INDEX

Explanations

proper nouns related to politics and individuals associated with them

oai_token-act-pair · gpt-3.5-turbo

mentions of specific individuals or entities within a context of accountability or newsworthiness

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

uckland

-0.83

ingen

-0.78

atively

-0.69

osion

-0.69

ayers

-0.68

falls

-0.68

oppy

-0.66

aying

-0.65

uring

-0.65

olation

-0.64

POSITIVE LOGITS

illac

0.90

rov

0.84

jet

0.75

SHIP

0.75

arov

0.73

 Stanton

0.72

WARE

0.72

mbol

0.72

meric

0.72

 GOODMAN

0.70

Activations Density 0.012%

proper nouns related to politics and individuals associated with them

mentions of specific individuals or entities within a context of accountability or newsworthiness

No Comments

No Known Activations