INDEX

Explanations

terms related to federal government and policies

oai_token-act-pair · gpt-3.5-turbo

references to the federal government and its policies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 whilst

-0.81

 Picks

-0.78

 Whilst

-0.75

comfort

-0.70

arse

-0.69

etter

-0.68

joy

-0.68

Sense

-0.66

(*

-0.66

alys

-0.66

POSITIVE LOGITS

 federal

3.46

Federal

2.85

 Federal

2.53

 feds

2.51

 federally

2.26

ederal

2.07

 government

1.64

 provincial

1.62

 municipal

1.56

 governmental

1.52

Activations Density 0.027%

terms related to federal government and policies

references to the federal government and its policies

No Comments

No Known Activations