INDEX

Explanations

references to specific entities such as companies, individuals, and products

oai_token-act-pair · gpt-3.5-turbo

special characters or symbols used in format or emphasis

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

omorphic

-0.67

 metic

-0.65

omorph

-0.65

åŃĲ

-0.64

 Asgard

-0.64

ä½ľ

-0.64

ª

-0.63

 PACK

-0.62

ãĥĦ

-0.61

 artif

-0.61

POSITIVE LOGITS

requires

0.74

votes

0.74

while

0.69

feat

0.68

were

0.67

drivers

0.67

peror

0.66

Rated

0.65

depending

0.65

tests

0.64

Activations Density 0.169%

references to specific entities such as companies, individuals, and products

special characters or symbols used in format or emphasis

No Comments

No Known Activations