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

ase

-0.82

pair

-0.74

ispers

-0.70

pleasant

-0.67

fro

-0.66

PI

-0.66

neutral

-0.65

yea

-0.65

atlantic

-0.62

ivalent

-0.62

POSITIVE LOGITS

ulla

0.85

ĪĴ

0.75

 Lenin

0.70

 charism

0.69

obar

0.68

senal

0.67

ogun

0.65

pta

0.65

 Laksh

0.64

Serv

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.