INDEX

Explanations

terms related to government, surveillance, media, and controversies

oai_token-act-pair · gpt-3.5-turbo

references to major institutions or organizations

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

 Thiel

-0.60

 unfocusedRange

-0.59

 Beir

-0.57

 Spiel

-0.55

 Kimmel

-0.54

 Beware

-0.53

zik

-0.53

 FEMA

-0.52

ctuary

-0.51

 Ignore

-0.51

POSITIVE LOGITS

 workforce

0.65

 wardrobe

0.62

Reviewer

0.61

ranch

0.60

 subsequ

0.58

 hairst

0.58

 endeavors

0.57

 asset

0.55

 behaviours

0.54

 jobs

0.54

Activations Density 0.974%

terms related to government, surveillance, media, and controversies

references to major institutions or organizations

No Comments

No Known Activations