INDEX

Explanations

phrases related to political figures

oai_token-act-pair · gpt-3.5-turbo

colons followed by narrative elements or quotations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

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

chy

-0.64

rawdownloadcloneembedreportprint

-0.63

 respawn

-0.61

NetMessage

-0.57

 gunshot

-0.57

 depress

-0.57

pell

-0.56

 fing

-0.56

 rather

-0.56

cape

-0.55

POSITIVE LOGITS

 Cosponsors

0.67

©¶æ

0.63

 Hearing

0.63

 Results

0.62

 Probe

0.61

ãĥĨ

0.60

 Adviser

0.60

 Countries

0.59

 Grip

0.59

 Critics

0.58

Activations Density 0.022%

phrases related to political figures

colons followed by narrative elements or quotations

No Comments

No Known Activations