INDEX

Explanations

people's names or mentions of political figures in specific contexts

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

the frequency of the word "politics" and related terms

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

-0.70

 duty

-0.66

wom

-0.65

duty

-0.63

flix

-0.58

CI

-0.57

lif

-0.56

eele

-0.56

tsy

-0.54

 Duty

-0.54

POSITIVE LOGITS

ess

0.82

arily

0.80

ames

0.74

essed

0.74

ourn

0.71

irs

0.71

ĸļ

0.70

ame

0.67

anz

0.67

udes

0.66

Activations Density 0.233%

people's names or mentions of political figures in specific contexts

the frequency of the word "politics" and related terms

No Comments

No Known Activations