INDEX

Explanations

questions and inquiries within the text

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

questions

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

 getLayout

-0.41

 ché

-0.41

 الصف

-0.41

 packages

-0.40

userdetails

-0.38

VPB

-0.38

 AssemblyVersion

-0.38

 Paglinawan

-0.38

 stabilization

-0.37

 signatures

-0.37

POSITIVE LOGITS

 question

0.72

 pregunta

0.64

 Frage

0.61

gunta

0.60

 pergunta

0.60

vraag

0.60

QUESTION

0.59

 pytanie

0.59

 frågan

0.59

Question

0.57

Activations Density 0.043%

questions and inquiries within the text

questions

No Comments

No Known Activations