INDEX

Explanations

phrases related to legal matters, specifically involving threats and reporting incidents to authorities

oai_token-act-pair · gpt-3.5-turbo

commas and phrases related to explanations or clarifications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 staggered

-0.64

lun

-0.61

oe

-0.59

efe

-0.58

isy

-0.57

 output

-0.57

 deliber

-0.57

istrate

-0.56

 Rober

-0.56

 buttocks

-0.56

POSITIVE LOGITS

prototype

0.81

SourceFile

0.80

 Flavoring

0.76

esm

0.75

 Devices

0.74

specific

0.74

ãĤ¼

0.73

Origin

0.72

ITS

0.72

Inc

0.71

Activations Density 0.490%

phrases related to legal matters, specifically involving threats and reporting incidents to authorities

commas and phrases related to explanations or clarifications

No Comments

No Known Activations