INDEX

Explanations

mentions of suspected individuals involved in various criminal activities

oai_token-act-pair · gpt-3.5-turbo

the word "suspected" in various contexts related to allegations or accusations

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

ammy

-0.86

learn

-0.83

reen

-0.82

psey

-0.77

tem

-0.76

ental

-0.75

ajo

-0.75

skill

-0.71

alach

-0.71

umbn

-0.70

POSITIVE LOGITS

 suspect

0.88

 suspects

0.77

 culprit

0.74

 mishand

0.74

 infring

0.74

 suspected

0.72

 Innocent

0.67

 offenders

0.66

FTP

0.66

 misuse

0.65

Activations Density 0.009%

mentions of suspected individuals involved in various criminal activities

the word "suspected" in various contexts related to allegations or accusations

No Comments

No Known Activations