INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĳå£«

-0.76

 Deer

-0.74

 Carbuncle

-0.71

fect

-0.71

Dak

-0.70

 Inquisitor

-0.67

 Amar

-0.66

 Dartmouth

-0.66

©¶æ¥µ

-0.65

 Saber

-0.65

POSITIVE LOGITS

wake

0.76

Greek

0.75

SUP

0.74

pill

0.73

Proof

0.72

overs

0.72

skin

0.70

BRE

0.69

SELECT

0.69

ovsky

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.