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

nesday

-0.80

ursions

-0.70

ewitness

-0.66

abase

-0.64

 minimized

-0.64

aneers

-0.62

acement

-0.61

 displacement

-0.61

ursion

-0.60

oggle

-0.60

POSITIVE LOGITS

HTTP

0.83

DoS

0.78

buck

0.77

Newsletter

0.76

#$

0.74

favorite

0.70

Â§

0.68

Bug

0.66

ICAN

0.65

CVE

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.