INDEX

Explanations

the word "missing" or related words

oai_token-act-pair · gpt-3.5-turbo

references to missing persons

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Jinping

-0.69

irin

-0.66

gard

-0.64

grow

-0.63

idal

-0.63

xual

-0.62

idium

-0.60

gary

-0.60

 DRAG

-0.58

agonist

-0.57

POSITIVE LOGITS

0.88

 limbs

0.84

 persons

0.83

 Pieces

0.82

pelled

0.82

minded

0.81

 Persons

0.79

 ingredient

0.72

ives

0.70

 Malaysia

0.68

Activations Density 0.033%

the word "missing" or related words

references to missing persons

No Comments

No Known Activations