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

Negative Logits

 hardness

-0.64

owsky

-0.60

 pleading

-0.60

sr

-0.60

alsa

-0.60

strength

-0.59

cs

-0.59

Ready

-0.57

 MAKE

-0.57

[+

-0.57

POSITIVE LOGITS

ilater

0.78

vernment

0.77

gae

0.74

iations

0.70

gans

0.70

ulators

0.67

undreds

0.67

cit

0.67

asionally

0.66

omaly

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.