INDEX

Explanations

adjectives describing intensity or importance

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

éŃĶ

-0.79

 Recall

-0.78

ãĥİ

-0.78

 Surely

-0.73

CLAIM

-0.72

 âī¡

-0.71

ãĤ®

-0.69

 UPDATE

-0.69

Flickr

-0.68

UPDATE

-0.68

POSITIVE LOGITS

0.85

 holistic

0.83

 mathemat

0.79

 intangible

0.77

 transitional

0.75

 qualitative

0.75

 punitive

0.74

 layered

0.72

 deterrent

0.72

 gradual

0.71

Activations Density 0.303%

adjectives describing intensity or importance

No Comments

No Known Activations