INDEX

Explanations

questions about information and actions related to a specific topic

oai_token-act-pair · gpt-3.5-turbo

inquiries and prompts related to understanding and discussing specific topics or issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

IFIED

-0.66

ocaust

-0.62

entary

-0.57

anza

-0.56

rontal

-0.56

 Lesbian

-0.56

Son

-0.56

POR

-0.56

uclear

-0.56

ILE

-0.55

POSITIVE LOGITS

 accordingly

0.99

 thereof

0.96

 thereto

0.94

soever

0.91

 pitfalls

0.89

 implications

0.89

 obstacles

0.82

 consequ

0.80

abouts

0.79

 therein

0.78

Activations Density 0.211%

questions about information and actions related to a specific topic

inquiries and prompts related to understanding and discussing specific topics or issues

No Comments

No Known Activations