INDEX

Explanations

names of celebrities and public figures

oai_token-act-pair · gpt-3.5-turbo

keywords and names associated with political figures and events

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

 Brune

-0.66

auga

-0.65

 foundations

-0.64

 pains

-0.63

 swayed

-0.60

 Laurie

-0.60

 Wake

-0.59

KH

-0.59

Kem

-0.58

 Peel

-0.56

POSITIVE LOGITS

Ö¼

0.99

ardless

0.90

odore

0.86

]=

0.80

odo

0.75

imil

0.74

usterity

0.73

rontal

0.73

],[

0.71

vernment

0.71

Activations Density 0.093%

names of celebrities and public figures

keywords and names associated with political figures and events

No Comments

No Known Activations