INDEX

Explanations

general phrases indicating diversity or variety

oai_token-act-pair · gpt-3.5-turbo

references to a variety of categories or classifications of things

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

edia

-0.90

eka

-0.77

aeper

-0.69

itness

-0.67

ixon

-0.67

orney

-0.65

ylan

-0.65

ahime

-0.65

opsis

-0.65

lite

-0.64

POSITIVE LOGITS

 imaginable

0.92

 goodies

0.83

 kinds

0.79

 nasty

0.73

hots

0.72

 combinations

0.71

 fancy

0.71

hell

0.70

 funky

0.70

 conceivable

0.70

Activations Density 0.038%

general phrases indicating diversity or variety

references to a variety of categories or classifications of things

No Comments

No Known Activations