INDEX

Explanations

specific references to categories, possibly in a technological or organizational context

oai_token-act-pair · gpt-3.5-turbo

references to categories or classifications

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

bats

-0.91

 Zimmer

-0.77

Ney

-0.63

kj

-0.63

gotten

-0.62

RIS

-0.62

 Tavern

-0.61

ippi

-0.60

Dek

-0.59

 Lauder

-0.59

POSITIVE LOGITS

naire

1.04

ifier

0.93

category

0.78

 categories

0.77

ifiers

0.76

oola

0.75

Category

0.73

rss

0.73

naires

0.72

ãģĨ

0.72

Activations Density 0.042%

specific references to categories, possibly in a technological or organizational context

references to categories or classifications

No Comments

No Known Activations