INDEX

Explanations

the word "whim" at varying degrees of importance or relevance

oai_token-act-pair · gpt-3.5-turbo

mentions of impulsive or arbitrary decisions

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

kson

-0.94

retty

-0.82

Ã©s

-0.78

sson

-0.74

mosp

-0.73

IAL

-0.72

ial

-0.72

itÃ©

-0.72

qua

-0.72

folios

-0.72

POSITIVE LOGITS

ewitness

0.72

 Minute

0.70

 diaper

0.69

Pages

0.68

 cavity

0.68

 Haram

0.64

ibur

0.63

 whim

0.63

 inertia

0.63

 hamstring

0.62

Activations Density 0.052%

the word "whim" at varying degrees of importance or relevance

mentions of impulsive or arbitrary decisions

No Comments

No Known Activations