INDEX

Explanations

phrases indicating essential information or knowledge

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate essential information or knowledge

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Plaint

-0.68

 popularity

-0.63

 indefinitely

-0.57

 Reconstruction

-0.57

 Defendants

-0.57

redits

-0.57

 likeness

-0.56

oliath

-0.56

ques

-0.55

Output

-0.55

POSITIVE LOGITS

 KNOW

0.89

know

0.88

Know

0.87

 Know

0.82

Lear

0.74

 know

0.72

isode

0.68

 checklist

0.66

 SHOULD

0.65

IDER

0.64

Activations Density 0.141%

phrases indicating essential information or knowledge

phrases that indicate essential information or knowledge

No Comments

No Known Activations