INDEX

Explanations

specific keywords related to politics or specific entities

oai_token-act-pair · gpt-3.5-turbo

references to social and political issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

rals

-0.73

ousands

-0.70

 Flavoring

-0.64

 Cosponsors

-0.63

rils

-0.62

juries

-0.59

ãĥł

-0.59

osures

-0.59

leys

-0.58

ructure

-0.58

POSITIVE LOGITS

speak

1.23

 territory

1.06

 incarn

0.96

 jargon

0.95

 shorthand

0.94

 meets

0.92

 slang

0.91

actly

0.84

 heaven

0.84

 reborn

0.83

Activations Density 0.408%

specific keywords related to politics or specific entities

references to social and political issues

No Comments

No Known Activations