INDEX

Explanations

phrases related to telling someone to stop doing something

oai_token-act-pair · gpt-3.5-turbo

phrases indicating emotional or confrontational expressions

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

 Cosponsors

-0.62

cially

-0.58

etheless

-0.57

sequently

-0.55

ocument

-0.52

uably

-0.51

APH

-0.50

uilt

-0.49

ocally

-0.49

orously

-0.48

POSITIVE LOGITS

 fucking

0.76

 fuckin

0.75

 shit

0.69

 goddamn

0.68

 daddy

0.65

 crap

0.64

!",

0.62

 damn

0.61

 freaking

0.60

 godd

0.59

Activations Density 1.744%

phrases related to telling someone to stop doing something

phrases indicating emotional or confrontational expressions

No Comments

No Known Activations