INDEX

Explanations

legal or governmental terms related to decision-making and approval processes

oai_token-act-pair · gpt-3.5-turbo

terms related to governmental and legislative processes

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

uni

-0.75

Atl

-0.75

auri

-0.74

 iceberg

-0.73

 intangible

-0.73

 decomp

-0.71

ãĤ¨ãĥ«

-0.69

Pacific

-0.66

Afee

-0.66

mul

-0.65

POSITIVE LOGITS

 unanimously

1.04

 recess

1.03

Senate

0.95

 appoint

0.93

 convened

0.93

 Majority

0.90

 Appropriations

0.90

 appointed

0.89

 hearings

0.89

 Speaker

0.89

Activations Density 0.534%

legal or governmental terms related to decision-making and approval processes

terms related to governmental and legislative processes

No Comments

No Known Activations