INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

respons

-0.66

æķ

-0.66

iesta

-0.66

inct

-0.65

priv

-0.64

icio

-0.64

dig

-0.62

full

-0.62

sac

-0.62

GC

-0.62

POSITIVE LOGITS

elist

0.73

ablishment

0.72

eele

0.72

 Masquerade

0.71

arial

0.69

 rall

0.69

hander

0.66

untled

0.64

 discont

0.64

ezvous

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.