INDEX

Explanations

instances of the word 'no' used strongly in a negative context or to negate an action

oai_token-act-pair · gpt-3.5-turbo

phrases indicating negation or absence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

RAFT

-0.84

mosp

-0.65

romy

-0.63

inese

-0.58

jet

-0.58

 WATCHED

-0.57

chev

-0.57

aleb

-0.56

ioch

-0.56

çĶ°

-0.56

POSITIVE LOGITS

xious

1.26

 longer

1.10

 matter

0.91

ct

0.87

 doubt

0.86

except

0.85

obs

0.84

 indication

0.84

 consolation

0.79

otrop

0.78

Activations Density 0.038%

instances of the word 'no' used strongly in a negative context or to negate an action

phrases indicating negation or absence

No Comments

No Known Activations