INDEX

Explanations

mentions or variations of the word "ring"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "ring."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ultz

-0.77

merce

-0.68

EMP

-0.65

ufact

-0.65

 Hemp

-0.64

ACP

-0.64

grown

-0.63

urses

-0.60

phies

-0.59

 transitioning

-0.58

POSITIVE LOGITS

bone

1.01

rings

0.96

naire

0.96

naires

0.95

er

0.91

ring

0.90

ettes

0.88

tone

0.86

tons

0.84

leader

0.83

Activations Density 0.014%

mentions or variations of the word "ring"

instances of the word "ring."

No Comments

No Known Activations