INDEX

Explanations

names or words associated with individuals

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names and locations

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

krit

-0.71

 Kislyak

-0.65

 tremend

-0.64

yip

-0.64

doi

-0.63

 counsel

-0.61

 indebted

-0.59

 Kenobi

-0.58

ensional

-0.58

 Bethesda

-0.58

POSITIVE LOGITS

onne

0.81

reau

0.80

odor

0.77

utical

0.77

ohydrate

0.76

ohyd

0.75

heid

0.73

inho

0.72

nect

0.72

ades

0.72

Activations Density 0.098%

names or words associated with individuals

proper nouns, particularly names and locations

No Comments

No Known Activations