INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 prest

-0.76

 openings

-0.71

Å¡

-0.66

mos

-0.64

wow

-0.61

 rake

-0.61

zek

-0.61

ãĥ¼ãĥĨãĤ£

-0.60

enza

-0.60

 hedge

-0.59

POSITIVE LOGITS

 Sakuya

0.75

mitter

0.72

 Lindsay

0.71

arah

0.71

foundland

0.70

llah

0.69

itability

0.69

 Codex

0.67

 Transgender

0.66

SPONSORED

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.