INDEX

Explanations

mentions of how people are treated by others

oai_token-act-pair · gpt-3.5-turbo

references to the concept of treatment or being treated 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 @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

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

aer

-0.72

azi

-0.71

audi

-0.68

sky

-0.63

direction

-0.61

sign

-0.60

 Origin

-0.60

adra

-0.59

Rae

-0.58

activated

-0.58

POSITIVE LOGITS

ttes

0.87

ricular

0.83

 treated

0.79

reatment

0.78

terson

0.76

iments

0.75

ivated

0.75

illance

0.74

 pione

0.73

htaking

0.72

Activations Density 0.019%

mentions of how people are treated by others

references to the concept of treatment or being treated in various contexts

No Comments

No Known Activations