INDEX

Explanations

mentions of individuals who have been convicted of certain crimes

oai_token-act-pair · gpt-3.5-turbo

mentions of individuals who have been convicted of crimes

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

mentation

-0.85

mented

-0.76

hare

-0.70

idge

-0.68

gm

-0.68

ophe

-0.67

 Gork

-0.66

edia

-0.66

aven

-0.65

aged

-0.63

POSITIVE LOGITS

 felon

1.06

 guilty

0.88

 convict

0.84

icts

0.82

 perjury

0.80

 unfocusedRange

0.80

llor

0.80

fel

0.75

 wrongdoing

0.73

 convicted

0.72

Activations Density 0.040%

mentions of individuals who have been convicted of certain crimes

mentions of individuals who have been convicted of crimes

No Comments

No Known Activations