INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

isf

-0.74

aminer

-0.69

 perm

-0.65

rab

-0.65

 regards

-0.59

 decomp

-0.58

 respons

-0.58

 Cind

-0.57

 faire

-0.57

selves

-0.56

POSITIVE LOGITS

zens

0.75

puter

0.74

Ö¼

0.73

Versions

0.69

 Franken

0.67

ãĤ»

0.66

ATER

0.66

stros

0.65

IENCE

0.65

ophon

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.