INDEX

Explanations

mentions of violence or intense conflict

oai_token-act-pair · gpt-3.5-turbo

references to violence or tragic events, particularly those described as "bloody."

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

agall

-0.95

PLIED

-0.90

gravity

-0.87

Demand

-0.84

ership

-0.84

iculty

-0.83

 athlet

-0.80

rador

-0.78

rait

-0.77

essor

-0.76

POSITIVE LOGITS

 clashes

0.81

 dice

0.78

 bits

0.77

 bloody

0.76

 scars

0.76

 aftermath

0.74

 scraps

0.74

 adolesc

0.73

 noses

0.72

 diarrhea

0.72

Activations Density 0.060%

mentions of violence or intense conflict

references to violence or tragic events, particularly those described as "bloody."

No Comments

No Known Activations