INDEX

Explanations

scientific and abstract terms used in research papers

oai_token-act-pair · gpt-3.5-turbo

sections or headings in a document, particularly those that denote structural elements like abstracts, corrections, and results

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

nuts

-0.90

nut

-0.81

athered

-0.78

 jewels

-0.75

 pains

-0.74

 stewards

-0.73

 bells

-0.73

bled

-0.72

efully

-0.71

wood

-0.70

POSITIVE LOGITS

Abstract

1.47

 Abstract

1.02

 References

0.90

Invalid

0.90

 Definitions

0.90

gdala

0.90

 Definition

0.83

 Background

0.82

 Summary

0.80

Summary

0.79

Activations Density 0.023%

scientific and abstract terms used in research papers

sections or headings in a document, particularly those that denote structural elements like abstracts, corrections, and results

No Comments

No Known Activations