INDEX

Explanations

phrases related to news or information dissemination

oai_token-act-pair · gpt-3.5-turbo

instances of the word "NEWS" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ium

-0.65

etr

-0.64

ec

-0.63

umed

-0.63

 standing

-0.63

 lasting

-0.63

 object

-0.62

 reservation

-0.62

 mine

-0.62

 setting

-0.61

POSITIVE LOGITS

 NEWS

4.10

NEWS

2.32

news

1.77

 News

1.65

News

1.62

 news

1.47

 INFORMATION

1.23

 Newsletter

1.19

 VIDEOS

1.18

 POLIT

1.17

Activations Density 0.013%

phrases related to news or information dissemination

instances of the word "NEWS" in various contexts

No Comments

No Known Activations