INDEX

Explanations

numerical ranges expressed in terms of quantity

oai_token-act-pair · gpt-3.5-turbo

numerical values and statistics related to counts and measurements

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

Reviewer

-0.96

quickShipAvailable

-0.67

 Emblem

-0.65

 Twain

-0.65

Limited

-0.63

 ãĤµ

-0.61

="#

-0.61

 Called

-0.61

 Shuttle

-0.60

Myth

-0.60

POSITIVE LOGITS

 extent

0.85

 levels

0.83

 lengths

0.82

 feet

0.78

uate

0.78

 ratio

0.73

 degree

0.72

venge

0.72

 inclusive

0.71

mph

0.71

Activations Density 0.135%

numerical ranges expressed in terms of quantity

numerical values and statistics related to counts and measurements

No Comments

No Known Activations