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

 Ended

-0.65

utra

-0.63

 breastfeeding

-0.63

 strained

-0.63

 tense

-0.62

 tempor

-0.61

ktop

-0.61

whe

-0.61

 vigil

-0.60

mil

-0.60

POSITIVE LOGITS

iously

0.70

assis

0.69

ayson

0.68

BMC

0.68

acies

0.67

UPS

0.66

 Norn

0.65

 DRAG

0.65

ansas

0.62

Sys

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.