INDEX

Explanations

adjectives and descriptions related to people and their behaviors

oai_token-act-pair · gpt-3.5-turbo

sociopolitical labels and identities

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

-0.68

 partName

-0.62

 GOODMAN

-0.61

tails

-0.58

iland

-0.57

ancies

-0.55

mentioned

-0.54

uyomi

-0.52

mares

-0.52

 waiver

-0.52

POSITIVE LOGITS

 inferior

0.70

 discipl

0.66

 underdog

0.65

 unworthy

0.65

 undermin

0.62

 achievable

0.61

 martyr

0.60

 deserving

0.60

 savior

0.59

 superior

0.58

Activations Density 0.656%

adjectives and descriptions related to people and their behaviors

sociopolitical labels and identities

No Comments

No Known Activations