INDEX

Explanations

phrases mentioning specific names or proper nouns, especially related to news reporting or media personalities

oai_token-act-pair · gpt-3.5-turbo

mentions of a specific individual, likely a prominent figure in media or politics

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

stood

-0.78

historic

-0.69

angers

-0.68

omorphic

-0.68

population

-0.66

ties

-0.66

 validation

-0.65

 virginity

-0.64

 relegation

-0.64

 stocking

-0.62

POSITIVE LOGITS

 GOODMAN

2.00

gdala

1.02

APH

0.93

ENN

0.84

AN

0.81

ONES

0.79

ENCY

0.79

NER

0.78

AMY

0.77

aku

0.77

Activations Density 0.005%

phrases mentioning specific names or proper nouns, especially related to news reporting or media personalities

mentions of a specific individual, likely a prominent figure in media or politics

No Comments

No Known Activations