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

Disc

-0.84

ibles

-0.76

wcs

-0.69

Stra

-0.68

 toile

-0.68

Foot

-0.67

Priv

-0.67

emin

-0.65

Doug

-0.65

Doc

-0.65

POSITIVE LOGITS

 signalling

0.74

istani

0.69

ageddon

0.69

bley

0.68

 dynam

0.65

 patrolling

0.65

 towed

0.65

ilst

0.64

 reigning

0.61

INGTON

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.