INDEX

Explanations

terms related to negative social behaviors or medical conditions

oai_token-act-pair · gpt-3.5-turbo

terms related to antisocial behavior and its incidence 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 @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

å£«

-0.98

loo

-0.85

cia

-0.81

aby

-0.78

aults

-0.76

peat

-0.74

atche

-0.74

ikan

-0.74

ilts

-0.73

lins

-0.73

POSITIVE LOGITS

 incidence

0.87

cipline

0.70

 INST

0.69

 è£ıè¦ļéĨĴ

0.68

 sidx

0.65

uated

0.64

oral

0.63

 Jude

0.63

uations

0.63

 Malfoy

0.63

Activations Density 0.025%

terms related to negative social behaviors or medical conditions

terms related to antisocial behavior and its incidence in various contexts

No Comments

No Known Activations