1: % paper by W. Taylor on tachyon analysis from Pade + perturbation theory

2: 

3: %  \documentstyle{article}

4: %\documentclass[draft,12pt]{JHEP}

5: \documentclass[12pt]{JHEP}

6: %\documentstyle{article}

7: \usepackage{epsfig}

8: 

9: \font\blackboard=msbm10 at 12pt

10: %\font\blackboard=msbm12

11: \font\blackboards=msbm7

12: \font\blackboardss=msbm5

13: \newfam\black

14: \textfont\black=\blackboard

15: \scriptfont\black=\blackboards

16: \scriptscriptfont\black=\blackboardss

17: \def\bb#1{{\fam\black\relax#1}}

18: \def\BZ{\bb Z}

19: 

20: %\textheight=9.2in

21: \textwidth=6.5in

22: \oddsidemargin 1in	% LEFT MARGIN FOR ODD PAGES (20)

23: \evensidemargin 1in	% LEFT MARGIN FOR EVEN PAGES (20)

24: %\headheight=0in

25: %\headsep=0in

26: \topmargin=0in

27: %\oddsidemargin=0in

28: %

29: %% abbreviations for latex commands

30: 

31: 

32: \newcommand{\NP}{{\rm Nucl.\ Phys.\ }}

33: \newcommand{\AP}{{\rm Ann.\ Phys.\ }}

34: \newcommand{\PL}{{\rm Phys.\ Lett.\ }}

35: \newcommand{\PR}{{\rm Phys.\ Rev.\ }}

36: \newcommand{\PRP}{{\rm Phys.\ Rep.\ }}

37: \newcommand{\CMP}{{\rm Comm.\ Math.\ Phys.\ }}

38: \newcommand{\MPL}{{\rm Mod.\ Phys.\ Lett.\ }}

39: \newcommand{\PRL}{{\rm Phys.\ Rev.\ Lett.\ }}

40: \newcommand{\IJMP}{{\rm Int.\ J.\ Mod.\ Phys.\ }}

41: 

42: \def\identity{{\rlap{1} \hskip 1.6pt \hbox{1}}}

43: \def\Tr{{\rm Tr}}

44: 

45: \def\math@note#1{\gdef\@eqnlabel{LAB: #1}}

46: %\draftmode

47: \title{A Perturbative Analysis of Tachyon Condensation}

48: \author{Washington Taylor\\

49: {Center for Theoretical Physics} \\

50: {MIT, Bldg.  6-308} \\

51: {Cambridge, MA 02139, U.S.A.} \\

52: {\tt wati@mit.edu}}

53: 

54: \abstract{Tachyon condensation in the open bosonic string is analyzed

55: using a perturbative expansion of the tachyon potential around the

56: unstable D25-brane vacuum.  Using the leading terms in the tachyon

57: potential, Pad\'e approximants can apparently give the energy of the

58: stable vacuum to arbitrarily good accuracy.  Level-truncation

59: approximations up to level 10 for the coefficients in the tachyon

60: potential are extrapolated to higher levels and used to find

61: approximants for the full potential.  At level 14 and above, the

62: resulting approximants give an energy less than -1 in units of the

63: D25-brane tension, in agreement with recent level-truncation results

64: by Gaiotto and Rastelli.  The extrapolated energy continues to

65: decrease below -1 until reaching a minimum near level 26, after which the

66: energy turns around and begins to approach -1 from below.  Within the

67: accuracy of this method, these results are completely consistent with

68: an energy which approaches -1 as the level of truncation is taken to

69: be arbitrarily large.}

70: 

71: \keywords{String field theory}

72: \preprint{MIT-CTP-3298, hep-th/0208149}

73: \begin{document}

74: 

75: %\let\old@label\label					% OLD LATEX COMMAND.

76: %\renewcommand{\label}[1]{%\name{ref-#1}%		% LABEL IN HYPER TAG.

77: %			 \old@label{#1}%		%

78: %			 \math@note{#1}%

79: %			 }

80: 

81: 

82: 

83: \baselineskip16pt

84: \parskip=4pt

85: 

86: %--------------------------------

87: \section{Background}

88: \label{sec:background}

89: 

90: Much effort has been expended over the last few years to confirm Sen's

91: conjectures \cite{Sen-universality} on tachyon condensation in the

92: open bosonic string (for reviews see \cite{reviews}).  Sen suggested

93: that the unstable open string vacuum in 26 dimensions should be

94: understood as containing a space-filling D25-brane.  Sen's first

95: conjecture states that the open bosonic string has a nontrivial

96: locally stable vacuum with energy $-T_{25} V$ relative to the unstable

97: perturbative open string vacuum, where $T_{25}$ is the tension of the

98: D25-brane, and $V$ is the volume of spacetime; this new stable vacuum

99: should correspond to the true closed string vacuum with no D-branes.

100: This conjecture has been tested by several authors using the approach

101: of truncating Witten's open string field theory \cite{Witten-SFT} to

102: include only fields below a certain total oscillator level.  In

103: \cite{ks-open} Kostelecky and Samuel used level truncation up to level

104: 4, including 10 fields.  These authors found that at low levels of

105: truncation Witten's theory has a stable vacuum solution, and that the

106: energy of this stable vacuum seems to converge as the level of

107: truncation of the theory is increased.  At that time, however,

108: D-branes had not yet been understood, and there was no natural way to

109: interpret the energy of this new vacuum.  In \cite{Sen-Zwiebach} Sen

110: and Zwiebach reconsidered this calculation in the light of Sen's

111: conjectures.  They found that at level 4, the vacuum energy of the

112: nontrivial solution of the level-truncated theory is 0.986 times the

113: energy $-E_0 =-T_{25}V$ predicted by Sen.  In \cite{Moeller-Taylor},

114: Moeller and Taylor extended this calculation to higher levels.  We

115: found that as the level of truncation is increased, the vacuum energy

116: continues to decrease, until at level 10 the vacuum energy reaches

117: $-0.9991\, E_0$.  These calculations seemed to indicate that level

118: truncation gives a systematic approximation to the full string field

119: theory, and that the vacuum energy computed in level truncation would

120: converge monotonically to $-E_0$.  The accuracy of this picture was

121: thrown into doubt when Gaiotto and Rastelli recently announced

122: \cite{Gaiotto-Rastelli} that when the level of truncation is increased

123: further, the vacuum energy {\it overshoots} $-E_0$.  They found that

124: at level 14 the vacuum energy becomes $-1.0002\, E_0$, and that the

125: energy continues to decrease monotonically until at level 18 the

126: vacuum energy is $-1.0005\, E_0$.

127: 

128: 

129: In this paper we use an alternative approach based on perturbation

130: theory to estimate the energy of the nontrivial vacuum of open string

131: field theory.  We use level truncation to determine the values of the

132: coefficients in a perturbative expansion of the effective

133: zero-momentum tachyon potential around the usual open string vacuum.

134: We then extrapolate these values to higher levels, and use the actual and

135: predicted values of the coefficients at various levels to compute Pad\'e

136: approximants for the effective potential.  The resulting Pad\'e

137: approximants seem to give highly reliable estimates for the vacuum

138: energy.  Furthermore, the Pad\'e approximants based on the extrapolated

139: perturbative coefficients give rise to energies which closely match

140: those found by Gaiotto and Rastelli.  After overshooting the desired

141: value of $-E_0$ for the energy, however, the predicted energies turn

142: around again near level 26, eventually returning close to the desired

143: value.  These calculations show that the results of Gaiotto and Rastelli

144: are fully compatible both with earlier computations of the

145: coefficients of the effective tachyon potential and with the

146: conjecture that the energy of the nontrivial vacuum in level-truncated

147: SFT asymptotically approaches the value predicted by Sen as the level

148: is taken arbitrarily large.

149: 

150: In this paper we work exclusively in Feynman-Siegel gauge.  Other

151: choices of gauge were explored in \cite{Ellwood-Taylor-gauge}; for

152: some other gauge choices, the level-truncated approximations to the

153: energy were found to go below $-E_0$ (sometimes at levels as low as $L

154: = 4$), and to be non-monotonic.  The results of this paper suggest

155: that this is in fact the generic behavior of level-truncated

156: approximations in any gauge; we expect that applying the methods of

157: this paper to the calculation in other gauges will give similar

158: results.

159: 

160: In Section 2 we discuss the calculation of coefficients in the

161: effective tachyon potential.  In Section 3 we describe the method of

162: Pad\'e approximants and their use in studying the vacuum energy of SFT.

163: In Section 4 we use Pad\'e approximants on the coefficients computed in

164: Section 2 and summarize our results.

165: 

166: %--------------------------------

167: \section{Perturbative tachyon potential}

168: \label{sec:perturbative}

169: 

170: The effective potential for the $p = 0$ tachyon field

171: $\phi$ is given by

172: \begin{eqnarray}

173: V (\phi)  & = & \sum_{n = 2}^{ \infty}  c_n (\kappa g)^{n-2} \phi^n

174: \label{eq:v}\\

175: & = & -\frac{1}{2}\phi^2 + (\kappa g) \phi^3 + c_4 (\kappa g)^2

176: \phi^4 + c_5 (\kappa g)^3 \phi^5 +\cdots\nonumber

177: \end{eqnarray}

178: where $\kappa=3^{7/2}/2^7 \approx 0.365$  and where

179: $c_4, c_5, \ldots$ are numerical constants.  Closed form expressions

180: for these coefficients were given in

181: \cite{ks-exact,Samuel-kn,WT-perturbative}.  These expressions cannot

182: be evaluated exactly, however, and must be approximated numerically.

183: Level truncation of Witten's string field theory gives an efficient

184: means of approximating these coefficients

185: \cite{ks-open,Moeller-Taylor,WT-perturbative}.

186: In \cite{Moeller-Taylor}, we computed the coefficients

187: up to $c_{60}$ at truncation levels up to (10, 20), meaning that all

188: fields up to level 10 and interactions up to total level 20 were

189: included.

190: 

191: For example, for $c_4$ successive level-truncation approximations give

192: \begin{eqnarray}

193: c_4^{[2]} & = & -\frac{34}{27} \approx -1.259259\\

194: c_4^{[4]} &  \approx & -1.472489 \nonumber\\

195: c_4^{[6]} &\approx & -1.556198 \nonumber\\

196: c_4^{[8]} &  \approx& -1.600425 \nonumber\\

197: c_4^{[10]} & \approx& -1.627694 \nonumber

198: \end{eqnarray}

199: where by $c_4^{[L]}$ we denote the approximation to $c_4$ in level

200: approximation $(L, 2L)$.  Note that throughout this paper we

201: systematically use level truncations $(L, 2L)$;  empirical

202: observations indicate that level $(L, 2L)$ approximations are

203: generally very close to level $(L, 3L)$ approximations, although

204: the former are easier to compute

205: \cite{Moeller-Taylor,Gaiotto-Rastelli}.   An analytic expression for

206: $c_4$ given in \cite{ks-exact} was numerically evaluated in

207: \cite{ks-open,WT-perturbative} and found to give 

208: \begin{equation}

209: c_4 \approx -1.742

210: \pm 0.001\,.

211: \label{eq:c4-mt}

212: \end{equation}

213: In \cite{WT-perturbative}, approximations to $c_4$ were computed using

214: oscillator level truncation of an exact expression written in terms

215: of an infinite matrix of Neumann coefficients.  These approximations

216: $c_4^{(L)}$ were computed up to oscillator level $L = 100$, and were

217: found to behave at large $L$ as

218: \begin{equation}

219: c_4^{(L)} \approx -1.742 + 0.80 \;L^{-1} +{\cal O} (L^{-2})\,.

220: \label{eq:c4-oa}

221: \end{equation}

222: 

223: To use the Pad\'e approximant method described in the next section, we

224: need an accurate approximation to the coefficients $c_{n}$ for $n$ up

225: to $26$ or so.  While the method of \cite{WT-perturbative} is much more

226: efficient for computing $c_n$ for small $n$ at high levels, at large

227: values of $n$ this approach becomes more difficult as the number of

228: diagrams which must be summed increases quickly with $n$.  Thus, in this paper

229: we will use the approximate coefficients $c_n^{[L]}$ computed in

230: \cite{Moeller-Taylor} up to level $L = 10$.  From the analysis of

231: \cite{WT-perturbative}, however, we expect that these approximate

232: coefficients will have corrections which can be expressed as a power

233: series in $1/L$.  This allows us to use the data for $L \leq 10$ to

234: predict the values of the approximate coefficients for $L > 10$.  For

235: example, using the values of $c_4^{[6]}, c_4^{[8]}, c_4^{[10]}$ from

236: (\ref{eq:c4-mt}) we can fit to a function of the form $a + b/L + c/L^2$ to get

237: \begin{equation}

238: c_4^{\{2, L\}} \approx-1.74226+0.59482 \, L^{-1}-0.10987\, L^{-2} \,.

239: \label{eq:c4-2}

240: \end{equation}

241: Using the value of $c_4^{[4]}$ in addition, and allowing a term of

242: order $L^{-3}$ gives

243: \begin{equation}

244: c_4^{\{3, L\}} \approx-1.74204+0.59213\, L^{-1}-0.09933 \,L^{-2} 

245: -0.01346 \, L^{-3}\,.

246: \label{eq:c4-3}

247: \end{equation}

248: Including $c_4^{[2]}$ and a term of order $L^{-4}$ gives

249: \begin{equation}

250: c_4^{\{4, L\}} \approx-1.74190+0.59013\, L^{-1}-0.08916 \,L^{-2} 

251: -0.03550 \, L^{-3} + 0.01718\, L^{-4}\,.

252: \label{eq:c4-4}

253: \end{equation}

254: The low order terms in these extrapolations seem to be converging

255: well, so we expect that (\ref{eq:c4-2}-\ref{eq:c4-4}) give

256: increasingly good estimates of the level-truncated approximations to

257: the coefficient $c_4$.  We have performed similar extrapolations for

258: the higher coefficients $c_n$ with similar results.  We will use these

259: extrapolations in Section 4 to estimate the value of the vacuum energy

260: in various level-truncated approximations.

261: 

262: 

263: 

264: %--------------------------------

265: \section{Pad\'e approximants}

266: \label{sec:Pade}

267: 

268: The power series expansion (\ref{eq:v}) of the tachyon effective

269: potential $V (\phi)$ has a finite radius of convergence of

270: approximately $\phi_r \approx 0.25/g$ \cite{Moeller-Taylor}.  This

271: finite radius of convergence arises due to the breakdown of

272: Feynman-Siegel gauge when $\phi$ approaches the critical value $\phi_c

273: \approx -0.25/g$ \cite{Ellwood-Taylor-gauge}.  At this point the

274: function $V (\phi)$ has a square root branch cut singularity.  The

275: nontrivial vacuum occurs near $\phi \approx 1/g$, so that the series

276: expansion of $V (\phi)$ is badly divergent near the true vacuum.

277: Thus, we cannot directly study the local minimum of the tachyon

278: effective potential by summing the perturbation series (\ref{eq:v}).

279: 

280: A powerful method for dealing with series expansions outside their

281: radius of convergence is given by the method of Pad\'e

282: approximants.  Given a power series

283: \begin{equation}

284: f (x) = a_0 + a_1 x + a_2x^2 + a_3x^3 + \cdots,

285: \label{eq:fx}

286: \end{equation}

287: the Pad\'e approximant $P^M_N (x)$ is the unique rational function

288: with numerator of degree $M$ and denominator of degree $N$ which

289: agrees with the power series expansion (\ref{eq:fx}) out to order $M +

290: N$

291: \begin{equation}

292: P^M_N (x) = \frac{\sum_{ i = 0}^{M} b_i x^i}{ \sum_{j = 0}^{N} d_j

293:   x^j}  = \sum_{k = 0}^{ N + M}  a_k x^k +{\cal O} (x^{N + M + 1})\,.

294: % \label{eq:}

295: \end{equation}

296: 

297: While the partial sums computed by truncating (\ref{eq:v}) at finite

298: order do not converge near the stable tachyon vacuum, these partial

299: sums may be used to compute Pad\'e approximants with much better

300: convergence behavior.  For example,

301: consider the approximate perturbative potential

302: given by truncating (\ref{eq:v}) at $\phi^K$ and using the level $2$

303: approximations $c_n^{[2]}$ to the coefficients.  

304: For $K = 4$, the resulting quartic 

305: \begin{equation}

306: -\frac{1}{2}\phi^2 + \kappa g \phi^3-\frac{34}{27}  (\kappa g) \phi^4

307: % \label{eq:}

308: \end{equation}

309: has no local minima, while the Pad\'e approximant

310: \begin{equation}

311: P^3_1 (\phi) =\frac{-\frac{1}{2}\phi^2 + \frac{10}{27} \kappa g \phi^3

312: }{1 + \frac{34}{27}  \kappa g \phi} 

313: % \label{eq:}

314: \end{equation}

315: does.

316: 

317: \FIGURE{

318: \epsfig{file=g1.eps,width=15cm}

319: \caption{\footnotesize Level 4 vacuum energy using Pad\'e approximants}

320: \label{f:Pade-4}

321: }

322: 

323: We have investigated the behavior of various families of Pad\'e

324: approximants using the coefficients of the tachyon effective potential

325: computed at low truncation levels.  We find that the family of

326: approximants $P^{n + 2}_n$ does very well at reproducing the effective

327: potential out to the nontrivial vacuum.  For example, at level 4 the

328: effective potential has a minimum with energy $-0.986403\, E_0$

329: \cite{Sen-Zwiebach,Moeller-Taylor}.  Using Pad\'e approximants on the level 4

330: coefficients we find that  the Pad\'e approximants $P^{n +

331: 2}_n$ give increasingly accurate estimates of vacuum energy, giving

332: for example

333: (in units of $E_0$)

334: \begin{eqnarray}

335: E^{12}_{ 10} & = &  -0.9864053 \nonumber\\

336: E^{13}_{11} & = &  -0.9864036\\

337: E^{14}_{12} & = & -0.9864033 \nonumber

338: \end{eqnarray}

339: The Pad\'e approximants converge up to the limit of numerical precision

340: possible from the coefficients we have used.  The energies arising from

341: using Pad\'e approximants $P^{n + 2}_n$ to estimate the vacuum energy at

342: level 4 are graphed in Figure~\ref{f:Pade-4}.  Performing a similar

343: analysis on the approximate coefficients at other levels of

344: truncation, we find that up to level 10, the Pad\'e approximant

345: $P^{14}_{12}$ comes within $10^{-6}$ of the exact value of the energy

346: at the minimum of the approximate effective potential.  We expect

347: similar behavior at higher levels.  We use this approximant in

348: our further analysis.

349: 

350: 

351: 

352: 

353: %--------------------------------

354: \section{Estimating the vacuum energy}

355: \label{sec:energy}

356: 

357: In Section 2 we described a method for estimating the coefficients

358: $c_n^{[L]}$ which would be computed using level truncation at levels

359: $L > 10$.  In section 3 we described a method for using coefficients

360: in the effective tachyon potential to estimate the vacuum energy.  We

361: can now combine these approaches to estimate the vacuum energy which

362: would be computed in a level $L$ truncation with $L > 10$.

363: 

364: \FIGURE{

365: \epsfig{file=g2.eps,width=15cm}

366: \caption{\footnotesize Extrapolating vacuum energy in level truncation

367: $L > 10$}

368: \label{f:extrapolate-L}

369: }

370: 

371: 

372: We have used the Pad\'e approximant $P^{14}_{12}$ to estimate the vacuum

373: energy which would arise from the tachyon potential given by each of

374: the three families of extrapolated coefficients $c_n^{\{m, L\}}$ given in

375: (\ref{eq:c4-2}-\ref{eq:c4-4}) for $m = 2, 3, 4$.  The resulting vacuum

376: energies are graphed in Figure~\ref{f:extrapolate-L} and compared to the

377: exact results of Gaiotto and Rastelli at levels 12-16

378: \cite{Gaiotto-Rastelli}.  As can be seen from the graph, while the

379: order $1/L^2$ extrapolation of the perturbative coefficients gives an

380: energy which overshoots the correct values, the order $1/L^4$

381: extrapolation is very close to the exact values.  For example, at

382: level 16 Gaiotto and Rastelli found

383: \begin{equation}

384: E^{[16]} = -1.0003678.

385: % \label{eq:}

386: \end{equation}

387: The order $1/L^4$ extrapolation gives

388: \begin{equation}

389: E^{\{4, 16\}} \approx -1.0003773

390: % \label{eq:}

391: \end{equation}

392: while the order $1/L^3$ and $1/L^2$ extrapolations give $E^{\{3, 16\}}

393: \approx -1.0004030$ and $E^{\{2, 16\}} \approx -1.0005007$

394: respectively.

395: 

396: Seeing that we can fairly accurately reproduce the values found

397: by Gaiotto and Rastelli for the energy in the level-truncated theory

398: at levels $L > 10$, it is of great interest to extrapolate to still

399: higher values of $L$.  We have done this for each of the three

400: extrapolations we are using.  In each case, we find that at a

401: particular level the extrapolated energy reaches a minimum value and

402: then starts to climb back towards -1.   Table~\ref{t:energies} summarizes

403: for each of the extrapolations we are using the level at which the

404: extrapolated energy reaches a minimum, the energy at the corresponding

405: minimum, and the energy as $L \rightarrow \infty$ \TABLE{

406: \begin{tabular}{ || c  || c | c | c ||}

407: \hline

408: \hline

409: Order &  $L$ at minimum $E$  & $E_{{\rm min}}$ & $E_{\infty} $\\

410: \hline

411: \hline

412: $1/L^2 $ & 138 &-1.001247 & -1.001239\\

413: \hline

414: $1/L^3 $ & 34 &-1.000761 & -1.000444\\

415: \hline

416: $1/L^4 $ & 28 &-1.000661 & -1.000159\\

417: \hline

418: \hline

419: \end{tabular}

420: 

421: \caption{\footnotesize Minimum and asymptotic energies for different

422:   extrapolations}

423: \label{t:energies}

424: }

425: The extrapolated energy is graphed as a function of $\ln L$ in

426: Figure~\ref{f:log-energy}.

427: 

428: \FIGURE{

429: \epsfig{file=g3.eps,width=15cm}

430: \caption{\footnotesize Extrapolated vacuum energy for large $L$}

431: \label{f:log-energy}

432: }

433: 

434: It is clear that these results are compatible both with the

435: observation of Gaiotto and Rastelli that the energy shoots below -1 at

436: level 14 and with the conjecture that as $L \rightarrow \infty$ the

437: energy approaches $-1$ in a controlled fashion.  It would be nice to

438: have a clearer theoretical understanding of why the individual

439: coefficients in the perturbative expansion of the tachyon effective

440: potential have corrections described by a power series in $1/L$ in

441: level truncation, as this feature was crucial in getting the

442: computations in this paper to work out.

443: 

444: 

445: 

446: 

447: \section*{Acknowledgements}

448: 

449: I would like to thank Leonardo Rastelli, Ashoke Sen, and Ilya Sigalov

450: for helpful discussions.  I would also like to thank the Aspen Center

451: for Physics, where this work was done.  The numerical computations

452: described in this work were done using {\it Mathematica}.  This work

453: was supported by the DOE through contract \#DE-FC02-94ER40818.

454: 

455: 

456: %%%%%%%%%%%%%%%%%%%%%%

457: 

458: 

459: \normalsize

460: %\newpage

461: 

462: \bibliographystyle{plain}

463: %\bibliography{papers}

464: 

465: \begin{thebibliography}{10}

466: 

467: \bibitem{Sen-universality} 

468: % lookup 9911116

469: A.~Sen,

470: ``Universality of the tachyon potential,''

471: JHEP {\bf 9912}, 027 (1999),

472: {\tt hep-th/9911116}.

473: 

474: 

475: \bibitem{reviews}

476: % lookup 0102085

477: K.~Ohmori,

478: ``A review on tachyon condensation in open string field theories,''

479: {\tt hep-th/0102085};

480: % lookup 0109182

481: P.~J.~De Smet,

482: ``Tachyon condensation: Calculations in string field theory,''

483: {\tt hep-th/0109182};

484: % lookup 0111208

485: I.~Y.~Arefeva, D.~M.~Belov, A.~A.~Giryavets, A.~S.~Koshelev and P.~B.~Medvedev,

486: ``Noncommutative field theories and (super)string field theories,''

487: {\tt hep-th/0111208};

488: W.\ Taylor and B.\ Zwiebach, ``2001 TASI lectures'', {\it to appear}.

489: 

490: \bibitem{Witten-SFT}

491: E.\ Witten, ``Non-commutative geometry and string field theory,'' \NP {\bf

492:   B268} 253 (1986).

493: 

494: \bibitem{ks-open}

495: V.\ A.\ Kostelecky and S.\ Samuel, ``On a nonperturbative vacuum for the open

496:   bosonic string,'' \NP {\bf B336} 263 (1990).

497: 

498: \bibitem{Sen-Zwiebach}

499: % lookup 9912249

500: A.~Sen and B.~Zwiebach,

501: ``Tachyon condensation in string field theory,''

502: JHEP {\bf 0003}, 002 (2000),

503: {\tt hep-th/9912249}.

504: 

505: 

506: 

507: \bibitem{Moeller-Taylor}

508: % lookup 0002237

509: N.~Moeller and W.~Taylor,

510: ``Level truncation and the tachyon in open bosonic string field theory,''

511: Nucl.\ Phys.\ B {\bf 583}, 105 (2000),

512: {\tt hep-th/0002237}.

513: 

514: \bibitem{Gaiotto-Rastelli}

515: D.\ Gaiotto and L.\ Rastelli,

516: ``Progress in open string field theory,''

517: Presentation by L.\ Rastelli at Strings 2002, Cambridge, England;

518: {\tt http://www.damtp.cam.ac.uk/strings02/avt/rastelli/}.

519: 

520: \bibitem{Ellwood-Taylor-gauge} 

521: I.\ Ellwood and W.\ Taylor,

522: ``Gauge invariance and tachyon condensation in open string field

523: theory,''

524: {\tt hep-th/0105156}

525: 

526: 

527: \bibitem{ks-exact}

528: V.\ A.\ Kostelecky and S.\ Samuel, ``The static tachyon potential in the open

529:   bosonic string,'' \PL {\bf B207} 169 (1988).

530: 

531: \bibitem{Samuel-kn}

532: R.\ Bluhm and S.\ Samuel, ``The off-shell Koba-Nielsen formula,'' \NP

533: {\bf B323} 337 (1989). 

534: 

535: \bibitem{WT-perturbative}

536: W.~Taylor,

537: ``Perturbative diagrams in string field theory,''

538: {\tt hep-th/0207132}.

539: 

540: % lookup

541: %--------------------------------

542: 

543: \end{thebibliography}

544: 

545: 

546: \end{document}

547: 

548: 

549: