INDEX

Explanations

phrases indicating an abundance or frequency of something

oai_token-act-pair · gpt-3.5-turbo

phrases indicating frequent or significant occurrences and the use of modifiers that describe quantity or extent

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

 Niet

-0.76

 horizont

-0.71

 cracked

-0.61

 nailed

-0.61

 exceeded

-0.61

Mania

-0.60

DN

-0.60

cheon

-0.59

rought

-0.59

ridor

-0.58

POSITIVE LOGITS

 noises

0.94

 decisions

0.90

roads

0.85

 arrangements

0.85

 adjustments

0.84

 noise

0.83

 mistake

0.83

 impression

0.81

 strides

0.81

sense

0.81

Activations Density 0.286%

phrases indicating an abundance or frequency of something

phrases indicating frequent or significant occurrences and the use of modifiers that describe quantity or extent

No Comments

No Known Activations