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

 Hide

-0.66

IRD

-0.61

organic

-0.61

HD

-0.61

 Isles

-0.61

tics

-0.60

 Radiant

-0.59

 Democr

-0.58

UI

-0.58

 Associates

-0.58

POSITIVE LOGITS

geist

0.77

ascript

0.75

cius

0.72

resso

0.69

abies

0.69

nesota

0.67

conservancy

0.66

ritch

0.65

okemon

0.65

agine

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.