INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

VERTISEMENT

-0.73

ategor

-0.71

Islam

-0.70

å£«

-0.69

 externally

-0.68

[+

-0.66

 ILCS

-0.65

achelor

-0.65

Reloaded

-0.64

ãģ®éŃĶ

-0.63

POSITIVE LOGITS

pport

0.81

emouth

0.78

icz

0.77

zy

0.77

kes

0.73

sonian

0.72

blers

0.70

sels

0.70

irie

0.69

raq

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.