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

helm

-1.01

cius

-0.83

imore

-0.77

fred

-0.76

cmp

-0.73

daq

-0.72

arily

-0.71

edom

-0.71

ampunk

-0.71

arist

-0.71

POSITIVE LOGITS

0.74

0.73

0.70

0.69

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.