INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 compr

-0.74

ĸļ

-0.68

senal

-0.68

 norm

-0.65

lication

-0.64

igation

-0.64

erest

-0.64

fore

-0.64

ient

-0.63

GROUND

-0.62

POSITIVE LOGITS

 cloaked

0.61

 scanners

0.58

handler

0.57

jam

0.57

 Titan

0.57

noticed

0.57

pta

0.57

 Senior

0.56

 whistlebl

0.56

 scrib

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.