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

inki

-0.77

otin

-0.76

 Logged

-0.75

kB

-0.73

 Saud

-0.71

 Astral

-0.70

 Bett

-0.70

rontal

-0.70

thodox

-0.69

 Tatt

-0.67

POSITIVE LOGITS

ives

0.71

stood

0.64

serv

0.61

gre

0.59

MIN

0.59

inery

0.58

IME

0.58

driver

0.58

arm

0.57

 ministries

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.