INDEX

Explanations

references to specific quantities or measurements

oai_token-act-pair · gpt-3.5-turbo

phrases related to locations or settings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ricanes

-0.63

 outl

-0.57

emb

-0.56

forts

-0.55

interstitial

-0.55

jit

-0.54

Fax

-0.54

sten

-0.53

bos

-0.53

 outsiders

-0.53

POSITIVE LOGITS

 existence

1.18

 sight

0.95

 circulation

0.92

 attendance

0.88

 contention

0.88

between

0.88

clusions

0.80

 chronological

0.78

existence

0.76

 succession

0.76

Activations Density 0.155%

references to specific quantities or measurements

phrases related to locations or settings

No Comments

No Known Activations