INDEX

Explanations

references to reporters or journalists

oai_token-act-pair · gpt-3.5-turbo

references to journalists or reporters

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

inence

-0.82

ateral

-0.78

ensions

-0.77

perm

-0.74

ersen

-0.74

aline

-0.71

ertodd

-0.71

emonic

-0.71

¥ŀ

-0.70

illard

-0.69

POSITIVE LOGITS

 reporter

0.91

 reporters

0.84

 COVER

0.75

ially

0.74

 Reporter

0.74

 journalists

0.73

 Journalists

0.73

 laureate

0.72

 interviewed

0.71

 Coverage

0.70

Activations Density 0.020%

references to reporters or journalists

references to journalists or reporters

No Comments

No Known Activations