INDEX

Explanations

names of political figures

oai_token-act-pair · gpt-3.5-turbo

specific geographical or cultural identifiers and names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 envy

-0.61

OPLE

-0.55

 kittens

-0.55

FACE

-0.54

staking

-0.53

 condem

-0.53

lished

-0.53

 notch

-0.52

 readiness

-0.50

 puzz

-0.50

POSITIVE LOGITS

oli

0.76

ema

0.75

ak

0.75

ich

0.74

oz

0.74

am

0.73

rad

0.72

ana

0.72

aj

0.71

ar

0.71

Activations Density 0.319%

names of political figures

specific geographical or cultural identifiers and names

No Comments

No Known Activations