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

atts

-0.73

 commands

-0.65

 privileges

-0.64

ango

-0.62

 prescribing

-0.62

 propositions

-0.61

 results

-0.61

 treatments

-0.60

isable

-0.60

 Norman

-0.59

POSITIVE LOGITS

ÃįÃį

0.84

Ŀ

0.82

culosis

0.80

ãĤ¨ãĥ«

0.77

deen

0.77

assis

0.73

phrine

0.72

 playthrough

0.71

eteenth

0.71

Eid

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.