INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Proud

-0.68

kay

-0.66

Redditor

-0.65

%);

-0.64

 Dangerous

-0.63

ailable

-0.63

 marked

-0.62

iannopoulos

-0.61

-0.60

 McKay

-0.60

POSITIVE LOGITS

 tempo

0.95

 stall

0.73

 situational

0.72

 Situation

0.70

 disapp

0.66

 corridors

0.65

 assum

0.65

 clipboard

0.64

 corridor

0.64

ophon

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.