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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 disclaim

-0.73

eworthy

-0.71

 Sabha

-0.71

 congratulate

-0.71

inary

-0.69

 RELEASE

-0.69

"],

-0.69

OLOG

-0.68

 regretted

-0.67

ldom

-0.66

POSITIVE LOGITS

urches

0.66

Ability

0.64

 poles

0.62

waters

0.62

 benches

0.61

tones

0.61

 Webb

0.60

Phill

0.60

atism

0.59

pert

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.