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

acca

-0.87

agan

-0.84

inning

-0.80

MIT

-0.78

oven

-0.74

rek

-0.74

acity

-0.74

asive

-0.73

acha

-0.73

rod

-0.72

POSITIVE LOGITS

Hut

0.71

çīĪ

0.69

 stove

0.64

 Sins

0.62

 Survivors

0.62

Riv

0.62

 census

0.62

 sleeper

0.62

 hots

0.61

 survivors

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.