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

wcsstore

-0.84

bley

-0.75

 reimbursement

-0.72

 bows

-0.67

 refurb

-0.67

 rematch

-0.66

cu

-0.65

 ..."

-0.65

 reinforcement

-0.65

 clasp

-0.65

POSITIVE LOGITS

IVE

0.77

Area

0.73

atech

0.72

ngth

0.71

sels

0.69

 Kumar

0.68

act

0.68

olicy

0.68

ema

0.68

wm

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.