INDEX

Explanations

names of people or organizations

oai_token-act-pair · gpt-3.5-turbo

proper nouns, specifically names of individuals and locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 womb

-0.76

 chem

-0.71

 tune

-0.69

 Hurricane

-0.62

USC

-0.57

 Mandatory

-0.57

 millennials

-0.57

 trapping

-0.57

 reins

-0.56

 semester

-0.56

POSITIVE LOGITS

itsch

0.96

arde

0.96

cott

0.92

arb

0.91

zinski

0.90

igl

0.90

assi

0.88

 apologised

0.88

ovich

0.86

chin

0.86

Activations Density 0.143%

names of people or organizations

proper nouns, specifically names of individuals and locations

No Comments

No Known Activations