INDEX

Explanations

words related to physical actions or movements

oai_token-act-pair · gpt-3.5-turbo

various letter combinations or fragments occurring 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

Osw

-0.65

 Robo

-0.62

 contrace

-0.60

pid

-0.59

 flares

-0.58

 disadvant

-0.57

 envy

-0.57

 prox

-0.56

 Rosenberg

-0.55

 Rossi

-0.55

POSITIVE LOGITS

auld

1.01

ï¸ı

0.92

ois

0.85

rences

0.84

Ã©

0.80

ifix

0.77

schild

0.76

stown

0.75

oul

0.75

ous

0.73

Activations Density 0.136%

words related to physical actions or movements

various letter combinations or fragments occurring in the text

No Comments

No Known Activations