INDEX

Explanations

phrases related to policy and politics

oai_token-act-pair · gpt-3.5-turbo

patterns of disagreement or contention in discussions

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

 abandon

-0.71

boro

-0.68

 herself

-0.63

 comple

-0.63

 swall

-0.62

iki

-0.61

 lifes

-0.61

odon

-0.60

fulness

-0.59

 asleep

-0.59

POSITIVE LOGITS

ccording

0.90

Until

0.90

Scroll

0.89

Unless

0.87

Ur

0.85

Unlike

0.84

Even

0.84

Indeed

0.84

âĢ¢âĢ¢

0.83

Asked

0.82

Activations Density 0.593%

phrases related to policy and politics

patterns of disagreement or contention in discussions

No Comments

No Known Activations