INDEX

Explanations

words related to physical violence or harm inflicted on individuals by others

oai_token-act-pair · gpt-3.5-turbo

terms related to various forms of violence and assault

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tions

-0.73

istries

-0.72

alities

-0.68

FK

-0.67

antz

-0.66

negie

-0.64

 formation

-0.64

rium

-0.63

Zone

-0.62

issue

-0.61

POSITIVE LOGITS

by

1.02

 merciless

0.91

nikov

0.79

 aback

0.78

 repeatedly

0.76

BY

0.75

by

0.74

 anew

0.74

 relentlessly

0.73

 violently

0.73

Activations Density 0.160%

words related to physical violence or harm inflicted on individuals by others

terms related to various forms of violence and assault

No Comments

No Known Activations