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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 istg

-0.66

 rockets

-0.65

 nonex

-0.64

 ions

-0.64

 tanks

-0.63

 rotating

-0.61

 expend

-0.61

 viable

-0.61

cru

-0.60

 nont

-0.60

POSITIVE LOGITS

 Seym

0.97

 Bronx

0.89

 redes

0.73

ãĥ¼ãĥ³

0.72

ulators

0.72

çīĪ

0.69

eport

0.69

abel

0.68

ilial

0.67

é¾į

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.