INDEX

Explanations

specific nouns related to different categories or entities

oai_token-act-pair · gpt-3.5-turbo

prominent nouns and terms related 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 @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

NPR

-0.54

 Faul

-0.53

 Keller

-0.53

 McDonnell

-0.50

CHAT

-0.49

 Strategic

-0.49

lua

-0.49

anwhile

-0.49

 CNBC

-0.48

 Cinem

-0.48

POSITIVE LOGITS

 nodd

0.66

 imaginable

0.59

 endors

0.59

ista

0.58

together

0.58

ailable

0.57

chev

0.56

0.55

 shenan

0.55

milo

0.55

Activations Density 2.213%

specific nouns related to different categories or entities

prominent nouns and terms related to categories or classifications

No Comments

No Known Activations