INDEX

Explanations

references to places or things named "Little."

oai_token-act-pair · gpt-3.5-turbo

the frequent mention of the word "Little."

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ãĥķãĤ©

-0.87

icular

-0.82

ktop

-0.79

ickr

-0.75

anwhile

-0.75

 conclud

-0.73

ulhu

-0.71

ulkan

-0.71

oÄŁ

-0.70

inates

-0.69

POSITIVE LOGITS

finger

0.92

woods

0.83

 Bird

0.80

 Sisters

0.79

 Miss

0.79

Bit

0.78

 Little

0.77

wood

0.77

 Brother

0.76

bird

0.75

Activations Density 0.011%

references to places or things named "Little."

the frequent mention of the word "Little."

No Known Activations