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

ritz

-0.77

corn

-0.75

abetic

-0.74

ersen

-0.73

fen

-0.73

iatrics

-0.70

ADRA

-0.68

iotics

-0.67

PLA

-0.66

iotic

-0.65

POSITIVE LOGITS

voy

0.89

 Voyager

0.77

sheet

0.71

ysc

0.69

LIA

0.64

 convict

0.63

ãĤ¨ãĥ«

0.61

 Keeper

0.60

cape

0.59

 convent

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.