INDEX

Explanations

language that implies adding something more or extra

oai_token-act-pair · gpt-3.5-turbo

references to extra or supplementary information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

-0.83

bows

-0.76

dog

-0.70

ESE

-0.70

abad

-0.67

generic

-0.66

 Â¯

-0.66

walk

-0.65

cture

-0.63

geist

-0.63

POSITIVE LOGITS

judicial

0.89

requisite

0.88

 installments

0.85

 complication

0.84

 avenues

0.81

 expense

0.81

 additions

0.81

 incentive

0.80

 additional

0.80

elaide

0.80

Activations Density 0.015%

language that implies adding something more or extra

references to extra or supplementary information

No Comments

No Known Activations