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

ictionary

-0.77

zin

-0.70

abus

-0.70

azo

-0.70

alan

-0.70

reading

-0.70

thread

-0.69

stellar

-0.69

forum

-0.67

advant

-0.66

POSITIVE LOGITS

 CARE

0.74

 Doodle

0.71

CD

0.71

cd

0.70

dB

0.68

FTWARE

0.67

::::::::

0.66

 euth

0.65

aughtered

0.65

cue

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.