INDEX

Explanations

phrases related to government agencies and organizations

oai_token-act-pair · gpt-3.5-turbo

references to governmental and organizational monitoring and regulation

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

 flakes

-0.58

ovsky

-0.56

 nausea

-0.54

ï¸

-0.53

illation

-0.53

 torped

-0.52

opian

-0.51

 itch

-0.51

 haunting

-0.50

 drown

-0.50

POSITIVE LOGITS

 Colleges

0.71

 Feder

0.66

 Universities

0.65

Administ

0.65

lishes

0.64

 Telecommunications

0.64

incial

0.62

 academia

0.62

 headquartered

0.62

 Publication

0.61

Activations Density 1.335%

phrases related to government agencies and organizations

references to governmental and organizational monitoring and regulation

No Comments

No Known Activations