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

DAQ

-0.86

SPONSORED

-0.83

 Chocobo

-0.77

CRIP

-0.76

CTV

-0.76

hift

-0.75

oran

-0.74

ï¿½ï¿½ï¿½ï¿½

-0.73

DNA

-0.73

Newsletter

-0.72

POSITIVE LOGITS

theless

0.73

cling

0.68

urated

0.66

cled

0.65

 renewal

0.63

piring

0.61

goers

0.59

 intermitt

0.58

 teased

0.58

 empt

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.