INDEX

Explanations

text describing actions or activities related to physical objects or locations

oai_token-act-pair · gpt-3.5-turbo

punctuations or delineators in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

coni

-0.60

æ©

-0.56

 suspic

-0.56

é¾

-0.54

 explan

-0.54

Reward

-0.54

 frontline

-0.54

 proble

-0.53

avorite

-0.53

aky

-0.52

POSITIVE LOGITS

huh

0.85

etc

0.83

 please

0.69

please

0.66

 partName

0.64

govtrack

0.64

 which

0.60

 plus

0.60

rete

0.59

eh

0.59

Activations Density 0.191%

text describing actions or activities related to physical objects or locations

punctuations or delineators in the text

No Comments

No Known Activations