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

 CLSID

-0.68

 grep

-0.64

 Controls

-0.63

Gray

-0.61

Fighting

-0.61

ilar

-0.59

Dialogue

-0.59

 bruising

-0.59

iltration

-0.58

Dial

-0.58

POSITIVE LOGITS

OOK

0.77

Nun

0.73

cher

0.70

 Sole

0.68

 Seal

0.66

LOCK

0.65

 Nieto

0.65

MER

0.65

HHS

0.63

umen

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.