INDEX

Explanations

phrases related to country names and organizations, particularly ones related to legal matters like patents and trademarks

oai_token-act-pair · gpt-3.5-turbo

references to specific geographic locations and associated organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

wake

-0.76

tumblr

-0.71

dylib

-0.70

gin

-0.68

enegger

-0.64

itness

-0.64

 hers

-0.63

drm

-0.62

wreck

-0.62

 heartbeat

-0.61

POSITIVE LOGITS

Tob

0.80

ãģª

0.76

 Trad

0.70

Nev

0.69

 Conditions

0.65

 Monteneg

0.64

abis

0.63

ĳå£«

0.62

glers

0.62

å°Ĩ

0.61

Activations Density 0.127%

phrases related to country names and organizations, particularly ones related to legal matters like patents and trademarks

references to specific geographic locations and associated organizations

No Comments

No Known Activations