INDEX

Explanations

financial terms, names of individuals often associated with politics or business, and specific brand names

oai_token-act-pair · gpt-3.5-turbo

references to political figures and entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ingly

-0.84

Ĥ¬

-0.84

runners

-0.79

itives

-0.76

etry

-0.76

£ı

-0.76

Ł

-0.76

bley

-0.74

runner

-0.74

bing

-0.74

POSITIVE LOGITS

atinum

0.72

 Carly

0.72

 Surgery

0.70

 insurer

0.69

 Ducks

0.66

XVI

0.66

 surgeons

0.65

 Blanc

0.64

bourg

0.63

 Clause

0.63

Activations Density 0.049%

financial terms, names of individuals often associated with politics or business, and specific brand names

references to political figures and entities

No Comments

No Known Activations