INDEX

Explanations

references to soccer teams or clubs by their abbreviated names

oai_token-act-pair · gpt-3.5-turbo

references to football clubs or teams

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

lihood

-0.76

Downloadha

-0.73

bilt

-0.67

mond

-0.62

bringer

-0.62

xual

-0.60

ously

-0.60

ysis

-0.59

civ

-0.59

cler

-0.58

POSITIVE LOGITS

RIC

0.84

 Cincinnati

0.81

ORE

0.79

VII

0.76

erning

0.76

Ĥª

0.74

AT

0.73

ocused

0.73

soc

0.72

 Edmonton

0.72

Activations Density 0.023%

references to soccer teams or clubs by their abbreviated names

references to football clubs or teams

No Comments

No Known Activations