INDEX

Explanations

phrases referring to a specific position or state

oai_token-act-pair · gpt-3.5-turbo

phrases indicating various states of being or circumstances

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Flavoring

-0.96

favorite

-0.68

origin

-0.66

 Dresden

-0.63

 Medal

-0.63

âĵĺ

-0.61

 antid

-0.61

avorite

-0.60

glomer

-0.60

 Brist

-0.59

POSITIVE LOGITS

WARE

0.72

Ĥİ

0.72

achy

0.70

 haste

0.69

ossession

0.68

 toile

0.65

nir

0.64

docker

0.63

 hurry

0.63

phabet

0.61

Activations Density 0.044%

phrases referring to a specific position or state

phrases indicating various states of being or circumstances

No Comments

No Known Activations