INDEX

Explanations

mentions of journalists and reporters, especially those involved in specific events or detained

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

references to journalists and reporting activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

 Defin

-0.78

eenth

-0.71

 rebound

-0.70

ware

-0.70

edIn

-0.69

uctions

-0.68

plan

-0.66

lucent

-0.65

swing

-0.65

ukong

-0.64

POSITIVE LOGITS

 covering

1.30

 interviewing

1.08

 Seymour

1.04

 camer

1.00

 Serge

0.96

 photograp

0.95

 reporter

0.91

 freel

0.91

 photographer

0.90

 journalists

0.90

Activations Density 0.124%

mentions of journalists and reporters, especially those involved in specific events or detained

references to journalists and reporting activities

No Comments

No Known Activations