INDEX

Explanations

locations or events related to violent incidents

oai_token-act-pair · gpt-3.5-turbo

words related to specific social groups and incidents involving violence or discrimination

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

ernels

-1.06

Frames

-0.90

osponsors

-0.89

ologies

-0.86

 equivalents

-0.85

 Zombies

-0.82

words

-0.82

brids

-0.81

 Feet

-0.79

 Languages

-0.79

POSITIVE LOGITS

 classmate

1.05

 staffer

0.98

 prostitute

0.98

 colleague

0.95

 employee

0.95

 protester

0.94

 woman

0.93

 helicopter

0.93

 clinic

0.90

 contractor

0.90

Activations Density 0.523%

locations or events related to violent incidents

words related to specific social groups and incidents involving violence or discrimination

No Comments

No Known Activations