INDEX

Explanations

phrases starting with "Before" or similar introductory statements

oai_token-act-pair · gpt-3.5-turbo

introductory phrases and contextual setups in discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĥ´

-0.85

ãĥĩ

-0.76

oad

-0.62

ãĥ³ãĤ¸

-0.61

numbered

-0.60

 Occupations

-0.60

é¾

-0.59

 goose

-0.58

far

-0.57

 reused

-0.56

POSITIVE LOGITS

 suffice

0.93

ookie

0.83

 acquaint

0.81

Ħ¢

0.79

LET

0.78

orthy

0.74

 caveats

0.74

 remember

0.74

 clarify

0.74

let

0.74

Activations Density 0.119%

phrases starting with "Before" or similar introductory statements

introductory phrases and contextual setups in discussions

No Comments

No Known Activations