INDEX

Explanations

information related to news articles or reports

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

words and phrases related to oversight and accountability in various contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

amaz

-0.79

AST

-0.78

AH

-0.74

omez

-0.73

 Colomb

-0.70

...]

-0.69

berto

-0.66

amic

-0.66

 Gamma

-0.63

amo

-0.63

POSITIVE LOGITS

EStreamFrame

0.69

orthy

0.64

 rall

0.64

fff

0.64

thodox

0.62

 aloud

0.60

existent

0.58

 fringe

0.57

edia

0.57

igel

0.57

Activations Density 0.155%

information related to news articles or reports

words and phrases related to oversight and accountability in various contexts

No Comments

No Known Activations