INDEX

Explanations

detailed statements outlining multiple reasons

oai_token-act-pair · gpt-3.5-turbo

content related to presenting reasons or justifications for a claim

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

heid

-0.65

film

-0.56

 Jagu

-0.55

entin

-0.54

 Falk

-0.53

 seiz

-0.53

 Presence

-0.53

istine

-0.52

 Resistance

-0.52

 settlers

-0.52

POSITIVE LOGITS

 myriad

1.17

 reasons

1.13

 several

0.94

 different

0.89

 fundamental

0.88

things

0.87

 ways

0.85

two

0.84

 numerous

0.84

 various

0.83

Activations Density 0.354%

detailed statements outlining multiple reasons

content related to presenting reasons or justifications for a claim

No Comments

No Known Activations