INDEX

Explanations

differentiating characteristics or qualities within a given context

oai_token-act-pair · gpt-3.5-turbo

references to varying categories and types

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

isman

-0.72

arton

-0.67

ItemTracker

-0.64

 earnest

-0.58

 irony

-0.58

gotten

-0.57

ridor

-0.57

 understatement

-0.56

 excess

-0.56

 patience

-0.55

POSITIVE LOGITS

 depending

1.45

paces

1.14

 differing

1.14

Different

1.07

 varying

0.98

 Different

0.96

 vying

0.95

styles

0.95

depending

0.95

 different

0.94

Activations Density 0.247%

differentiating characteristics or qualities within a given context

references to varying categories and types

No Comments

No Known Activations