INDEX

Explanations

sentences related to personal opinions and reflections

oai_token-act-pair · gpt-3.5-turbo

emotional expressions and opinions about relationships

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

Not in Any Lists

No Comments

Negative Logits

eatured

-0.51

urst

-0.50

respective

-0.48

 incumbent

-0.47

è£ħ

-0.47

atories

-0.46

 controvers

-0.46

odge

-0.45

idespread

-0.45

geoning

-0.45

POSITIVE LOGITS

 fuckin

0.87

 fucking

0.85

 goddamn

0.74

..."

0.70

 stupid

0.66

 fucked

0.66

 bitch

0.66

 shitty

0.66

 shit

0.64

 godd

0.63

Activations Density 3.005%

sentences related to personal opinions and reflections

emotional expressions and opinions about relationships

No Comments

No Known Activations