INDEX

Explanations

specific instances of the word "this" followed by another word or phrase

oai_token-act-pair · gpt-3.5-turbo

references to reading specific articles, posts, or communications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

roxy

-0.79

hesda

-0.72

aires

-0.71

ichick

-0.71

asia

-0.70

akable

-0.68

omaly

-0.67

ongevity

-0.66

arent

-0.66

otics

-0.66

POSITIVE LOGITS

 aloud

1.44

 excerpts

0.93

 papers

0.87

 instructions

0.84

 transcript

0.84

 passages

0.84

é¾įå¥ĳå£«

0.82

DragonMagazine

0.82

 reviews

0.81

 blogs

0.81

Activations Density 0.140%

specific instances of the word "this" followed by another word or phrase

references to reading specific articles, posts, or communications

No Comments

No Known Activations