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

Round

-0.63

shr

-0.61

forth

-0.61

rounder

-0.61

SN

-0.60

KR

-0.60

Orig

-0.59

mentioned

-0.59

letter

-0.58

 testing

-0.58

POSITIVE LOGITS

 Thief

0.77

esan

0.75

 âĶľ

0.73

gdala

0.73

-+-+

0.72

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

0.70

 Cumber

0.70

 Butterfly

0.70

atown

0.69

azaki

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.