INDEX

Explanations

government and policy-related terms and entities

oai_token-act-pair · gpt-3.5-turbo

terms related to governmental and organizational entities involved in health and social services

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

milo

-0.71

uckland

-0.71

buquerque

-0.62

cair

-0.61

onyms

-0.59

matically

-0.59

rain

-0.59

 captcha

-0.59

 ç¥ŀ

-0.58

azeera

-0.57

POSITIVE LOGITS

 alike

0.93

 Sciences

0.81

 Affairs

0.80

 Productions

0.79

 Club

0.78

 Association

0.77

 Services

0.77

 Development

0.76

 Products

0.75

 Center

0.73

Activations Density 0.160%

government and policy-related terms and entities

terms related to governmental and organizational entities involved in health and social services

No Comments

No Known Activations