INDEX

Explanations

numerical quantities

oai_token-act-pair · gpt-3.5-turbo

references to numerical limits or capacities

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

bane

-0.84

FORE

-0.84

 Guys

-0.75

perse

-0.75

çīĪ

-0.72

ãĥīãĥ©

-0.71

neys

-0.71

Aki

-0.70

rums

-0.70

Interstitial

-0.69

POSITIVE LOGITS

 amount

1.04

 allowable

1.04

 extent

0.93

nesota

0.90

 likelihood

0.88

egal

0.85

imize

0.83

 maximum

0.82

 punishment

0.80

ionage

0.79

Activations Density 0.011%

numerical quantities

references to numerical limits or capacities

No Comments

No Known Activations