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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ymes

-0.69

icken

-0.68

ploy

-0.66

 jumper

-0.63

Poc

-0.62

 downt

-0.61

ilk

-0.61

 Airl

-0.60

Dew

-0.60

 Weld

-0.59

POSITIVE LOGITS

ibal

0.90

aito

0.78

 caut

0.77

interstitial

0.75

acio

0.69

 suspic

0.68

Viz

0.67

aughter

0.65

osuke

0.63

burst

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.