INDEX

Explanations

phrases implying the truth or validity of a statement

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate contrasts or negations of statements

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ngth

-0.73

 freezes

-0.70

umbnails

-0.69

alin

-0.63

 clone

-0.61

 Scores

-0.61

usercontent

-0.59

enaries

-0.59

ription

-0.58

 hail

-0.56

POSITIVE LOGITS

 true

1.46

true

1.40

 happening

1.29

 untrue

1.15

false

0.99

 TRUE

0.99

 happen

0.98

 occurring

0.98

 possible

0.93

 achievable

0.92

Activations Density 0.162%

phrases implying the truth or validity of a statement

phrases that indicate contrasts or negations of statements

No Comments

No Known Activations