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

EStreamFrame

-0.89

reci

-0.72

ner

-0.69

croft

-0.66

 Arri

-0.66

Agg

-0.66

 heels

-0.66

gg

-0.65

arms

-0.65

 plun

-0.64

POSITIVE LOGITS

ALS

0.75

 fortunately

0.66

Ùħ

0.66

 contradiction

0.65

 impossibility

0.65

 Khalid

0.64

theless

0.62

 pine

0.62

 furthermore

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.