INDEX

Explanations

numerical expressions denoting distances or amounts

oai_token-act-pair · gpt-3.5-turbo

quantitative measures and statistics related to distance, time, and population

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

 Doodle

-0.68

istered

-0.66

 notwithstanding

-0.65

dale

-0.63

asta

-0.63

bra

-0.63

stre

-0.62

 Crimson

-0.62

 Slime

-0.61

puter

-0.60

POSITIVE LOGITS

dozen

0.81

cms

0.80

 thous

0.78

 dozen

0.74

 hours

0.72

hrs

0.71

 percent

0.70

 externalToEVAOnly

0.69

 pounds

0.69

OTAL

0.68

Activations Density 0.153%

numerical expressions denoting distances or amounts

quantitative measures and statistics related to distance, time, and population

No Comments

No Known Activations