INDEX

Explanations

numerical values and quantities mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

quantitative data related to losses or casualties

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

afety

-0.73

ulus

-0.68

unal

-0.66

otle

-0.62

arth

-0.61

 Solomon

-0.60

yo

-0.60

john

-0.59

sight

-0.58

ings

-0.57

POSITIVE LOGITS

CVE

0.62

aez

0.60

 bike

0.59

lopp

0.58

allery

0.57

 diff

0.56

Merit

0.56

EStream

0.56

hep

0.55

Ð»

0.54

Activations Density 0.304%

numerical values and quantities mentioned in the text

quantitative data related to losses or casualties

No Comments

No Known Activations