INDEX

Explanations

phrases related to different versions or options of something

oai_token-act-pair · gpt-3.5-turbo

specific technical and political terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Interstitial

-0.72

YING

-0.68

 Leban

-0.66

NetMessage

-0.65

steamapps

-0.64

 Alive

-0.64

eping

-0.60

nown

-0.60

cision

-0.60

cffffcc

-0.60

POSITIVE LOGITS

soever

0.98

 corresponds

0.85

 they

0.81

yip

0.75

 constitutes

0.73

he

0.72

 qualifies

0.71

we

0.70

you

0.69

 determines

0.68

Activations Density 0.097%

phrases related to different versions or options of something

specific technical and political terminology

No Comments

No Known Activations