INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

pheus

-0.79

igate

-0.79

anchester

-0.73

andise

-0.73

earable

-0.73

mob

-0.72

rx

-0.72

laws

-0.72

essage

-0.71

veh

-0.70

POSITIVE LOGITS

 underest

0.79

 Mist

0.61

soy

0.60

 coff

0.59

Lear

0.59

hiro

0.59

STL

0.58

Tang

0.58

 Cold

0.57

LCS

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.