INDEX

Explanations

sections or headers in a formal document that indicate important notes, proofs, or remarks

oai_token-act-pair · gpt-4o-mini Triggered by @bot

section headers or labels

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 betweenstory

-0.56

endregion

-0.48

脚注の使い方

-0.43

 nahilalakip

-0.43

发表于

-0.42

 AssemblyProduct

-0.40

'{@

-0.39

CharArray

-0.39

ectoria

-0.38

目に

-0.38

POSITIVE LOGITS

 purpoſe

0.47

 ſtre

0.46

 ſtand

0.46

 мәкал

0.45

 neceff

0.44

 neceſſ

0.44

 skjø

0.43

 poffible

0.43

arangay

0.42

 facteurs

0.42

Activations Density 0.028%

sections or headers in a formal document that indicate important notes, proofs, or remarks

section headers or labels

No Comments

No Known Activations