INDEX

Explanations

adjectives related to violence or conflict

oai_token-act-pair · gpt-3.5-turbo

references to violence or gore

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

uve

-0.86

PLIED

-0.82

Demand

-0.81

later

-0.80

agall

-0.78

Rate

-0.75

iculty

-0.74

ainer

-0.74

conservancy

-0.74

ering

-0.73

POSITIVE LOGITS

 noses

0.85

 wounds

0.83

 slaughter

0.79

 retribution

0.76

 scars

0.76

 clashes

0.75

 bloody

0.75

 revenge

0.75

 diarrhea

0.73

 wounding

0.73

Activations Density 0.053%

adjectives related to violence or conflict

references to violence or gore

No Comments

No Known Activations