INDEX

Explanations

references or citations in text

oai_token-act-pair · gpt-3.5-turbo

references to a specific source or authority within the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 incorpor

-0.74

 Enix

-0.73

ufact

-0.68

oro

-0.68

 scen

-0.67

?)

-0.66

ĺħ

-0.66

Luk

-0.65

catentry

-0.64

Klu

-0.63

POSITIVE LOGITS

 Reference

0.95

aez

0.67

fighting

0.66

Detect

0.65

cible

0.65

Category

0.64

hy

0.64

itcher

0.63

protected

0.63

help

0.62

Activations Density 0.000%

references or citations in text

references to a specific source or authority within the text

No Comments

No Known Activations