INDEX

Explanations

phrases discussing performance evaluations or critiques

oai_token-act-pair · gpt-3.5-turbo

criticisms related to performance and consistency in a competitive context

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

artisan

-0.88

idon

-0.82

ESE

-0.79

 democrat

-0.73

ACP

-0.73

 Inform

-0.72

toc

-0.72

velt

-0.70

 GOODMAN

-0.68

omers

-0.68

POSITIVE LOGITS

 inconsistency

1.49

 turnovers

1.43

 inconsistent

1.42

 inacc

1.30

 inability

1.28

 deficiencies

1.26

 sloppy

1.25

 injuries

1.24

 inconsistencies

1.24

 struggles

1.22

Activations Density 0.454%

phrases discussing performance evaluations or critiques

criticisms related to performance and consistency in a competitive context

No Comments

No Known Activations