INDEX

Explanations

phrases related to legal actions or classifications

oai_token-act-pair · gpt-3.5-turbo

specific terms or phrases related to role-playing games (RPGs)

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Rodrigo

-0.70

 Kern

-0.62

 Cree

-0.61

 Wake

-0.59

 Hutchinson

-0.59

 Salem

-0.57

 Noon

-0.57

nesday

-0.57

olphin

-0.56

imar

-0.56

POSITIVE LOGITS

conscious

0.99

appropriate

0.99

specific

0.98

appropri

0.96

checking

0.96

changing

0.95

definition

0.89

relation

0.89

change

0.88

classes

0.86

Activations Density 0.060%

phrases related to legal actions or classifications

specific terms or phrases related to role-playing games (RPGs)

No Comments

No Known Activations