INDEX

Explanations

instructions and recommendations related to actions or precautions

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

tips for advice

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

тельству

-0.48

ofern

-0.45

 gainera

-0.41

发表于

-0.39

 initData

-0.38

 mieste

-0.38

rensa

-0.38

 besz

-0.38

 pihak

-0.37

ător

-0.36

POSITIVE LOGITS

PointerException

0.44

 advice

0.43

 recipe

0.40

 tips

0.40

maske

0.40

TabStop

0.40

braucher

0.40

neapolis

0.39

OGND

0.39

 Dinge

0.39

Activations Density 0.031%

instructions and recommendations related to actions or precautions

tips for advice

No Comments

No Known Activations