INDEX

Explanations

phrases expressing strong opinions or emotions, especially negative ones

oai_token-act-pair · gpt-3.5-turbo

strong emotional or explicit language

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

cember

-0.58

 compounded

-0.56

inently

-0.55

 strikingly

-0.53

 sighting

-0.53

 resid

-0.52

xtap

-0.50

 converted

-0.49

 mosqu

-0.49

 estimated

-0.49

POSITIVE LOGITS

 ASAP

0.69

 Yourself

0.68

 anymore

0.66

 oneself

0.64

 yourselves

0.62

dirty

0.61

dress

0.58

 responsibly

0.57

 dirty

0.57

 damn

0.56

Activations Density 0.786%

phrases expressing strong opinions or emotions, especially negative ones

strong emotional or explicit language

No Comments

No Known Activations