INDEX

Explanations

technical terms or phrases related to legal, policy, and enforcement contexts

oai_token-act-pair · gpt-3.5-turbo

legal references and restrictions

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

 mirac

-0.79

 hopped

-0.73

 gorgeous

-0.72

 Sammy

-0.71

 crammed

-0.71

 beating

-0.71

 amazed

-0.70

 ruined

-0.69

 wrecked

-0.69

 towering

-0.68

POSITIVE LOGITS

Applic

1.17

 embodiments

1.13

 Applicant

1.09

Examples

1.09

shall

1.08

 Procedures

1.05

 Examples

1.05

 preferably

1.04

Â§

1.04

 Applic

1.00

Activations Density 0.670%

technical terms or phrases related to legal, policy, and enforcement contexts

legal references and restrictions

No Comments

No Known Activations