INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

isations

-0.70

ciation

-0.69

æ©Ł

-0.68

mins

-0.66

lists

-0.65

ayers

-0.65

ysis

-0.64

reads

-0.63

ãĥ³

-0.63

odes

-0.63

POSITIVE LOGITS

userc

0.73

Interstitial

0.70

 Workshop

0.68

nown

0.68

senal

0.63

fen

0.62

pez

0.62

inently

0.59

Pru

0.58

ategory

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.