INDEX

Explanations

mentions of governmental or political terms like "term" and "office" as well as verbs related to completing or running for office

oai_token-act-pair · gpt-3.5-turbo

references to political terms and durations of office

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

yout

-0.68

anooga

-0.66

acus

-0.66

 slightest

-0.66

 visceral

-0.62

ĳå£«

-0.59

 primal

-0.59

Obj

-0.58

 ingred

-0.58

 sophistication

-0.57

POSITIVE LOGITS

span

0.92

lasting

0.91

 ending

0.85

 lasting

0.85

 extending

0.84

 expired

0.83

long

0.83

 spanning

0.83

 consecut

0.81

 leng

0.80

Activations Density 0.128%

mentions of governmental or political terms like "term" and "office" as well as verbs related to completing or running for office

references to political terms and durations of office

No Comments

No Known Activations