INDEX

Explanations

mentions of the word "Hollywood" and related terms

oai_token-act-pair · gpt-3.5-turbo

references to Hollywood

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ktop

-0.91

arity

-0.90

theless

-0.79

İĭ

-0.72

haps

-0.71

etheless

-0.70

pron

-0.69

Ü

-0.69

eeper

-0.68

odic

-0.68

POSITIVE LOGITS

 Hills

1.04

 studios

1.01

 Reporter

1.00

 Hollywood

0.98

 Boulevard

0.98

 Studios

0.95

 celebrities

0.84

 Blvd

0.83

Film

0.83

mog

0.82

Activations Density 0.013%

mentions of the word "Hollywood" and related terms

references to Hollywood

No Comments

No Known Activations