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

Goal

-0.68

 Username

-0.68

ciation

-0.67

Attempts

-0.67

ADVERTISEMENT

-0.66

success

-0.66

Results

-0.64

Filename

-0.63

Forward

-0.63

Building

-0.63

POSITIVE LOGITS

ulkan

0.84

TPP

0.78

ttes

0.77

theless

0.76

gans

0.75

kees

0.71

apers

0.70

pots

0.69

afort

0.69

enf

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.