INDEX

Explanations

mentions of political figures, specifically senators

oai_token-act-pair · gpt-3.5-turbo

mentions of senators

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

order

-0.69

aque

-0.65

OPER

-0.64

MER

-0.64

 leash

-0.63

ãĥ¼ãĥ³

-0.63

 unpredictable

-0.62

 manual

-0.62

Company

-0.61

PER

-0.60

POSITIVE LOGITS

iors

1.02

hips

0.97

hip

0.94

eca

0.92

ileaks

0.92

 senators

0.87

 Senators

0.85

pai

0.81

 Dianne

0.74

etsk

0.73

Activations Density 0.013%

mentions of political figures, specifically senators

mentions of senators

No Comments

No Known Activations