INDEX

Explanations

words related to coding and programming, particularly in JavaScript

oai_token-act-pair · gpt-3.5-turbo

proper nouns and branding elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

govtrack

-0.64

rontal

-0.60

worldly

-0.59

inarily

-0.59

ciplinary

-0.56

bably

-0.56

incial

-0.55

 departures

-0.54

arbon

-0.53

ctions

-0.53

POSITIVE LOGITS

jong

0.54

 herself

0.54

igans

0.53

agher

0.51

Ame

0.51

AppData

0.51

 Sour

0.51

Shell

0.50

 Hemp

0.49

Gib

0.48

Activations Density 1.005%

words related to coding and programming, particularly in JavaScript

proper nouns and branding elements

No Comments

No Known Activations