INDEX

Explanations

comparative statistics related to rankings or levels

oai_token-act-pair · gpt-3.5-turbo

comparative phrases indicating extremes or significant numerical values

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

unes

-0.79

ertain

-0.74

ribut

-0.73

must

-0.66

Ļ

-0.65

ģĸ

-0.64

isSpecial

-0.63

rex

-0.62

Sport

-0.62

another

-0.62

POSITIVE LOGITS

 overall

0.88

 tendency

0.84

 propensity

0.83

 longevity

0.82

 beard

0.80

 amount

0.80

 reputation

0.79

 mileage

0.79

 versatility

0.78

 clout

0.77

Activations Density 0.138%

comparative statistics related to rankings or levels

comparative phrases indicating extremes or significant numerical values

No Comments

No Known Activations