INDEX

Explanations

communication interactions involving interviewing, speaking, and asking questions

oai_token-act-pair · gpt-3.5-turbo

dialogue or conversational phrases

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

)</

-0.80

estate

-0.72

FILE

-0.65

ruction

-0.65

interstitial

-0.64

 Himself

-0.63

-0.61

window

-0.61

Whatever

-0.60

 funer

-0.60

POSITIVE LOGITS

 extensively

0.73

 HuffPost

0.72

 empir

0.66

 qualitative

0.66

 testers

0.63

leneck

0.62

 firsthand

0.62

 commenter

0.62

 ourselves

0.62

 spoiler

0.61

Activations Density 0.491%

communication interactions involving interviewing, speaking, and asking questions

dialogue or conversational phrases

No Comments

No Known Activations