INDEX

Explanations

important or significant terms, potentially related to headings or section titles in a document

oai_token-act-pair · gpt-3.5-turbo

sections of text that contain definitions or related academic content

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hement

-0.74

 honoured

-0.72

 neighb

-0.71

userc

-0.69

 undermin

-0.69

ktop

-0.69

 slam

-0.69

 answ

-0.68

 princ

-0.67

 enthusi

-0.67

POSITIVE LOGITS

References

1.07

Trivia

0.92

Previously

0.90

Gallery

0.89

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

0.87

Production

0.86

Associated

0.86

 âĨĳ

0.85

Currently

0.84

âĨĳ

0.83

Activations Density 0.162%

important or significant terms, potentially related to headings or section titles in a document

sections of text that contain definitions or related academic content

No Comments

No Known Activations