INDEX

Explanations

structured and procedural elements in programming or documentation

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

answer text

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

 يتيمه

-0.93

 betweenstory

-0.76

IsContent

-0.75



-0.74

 المعيارى

-0.70

delwed

-0.69

windowFixed

-0.68

adaptiveStyles

-0.66

 ब्रेकडाउन

-0.66

󠁣

-0.66

POSITIVE LOGITS

<eos>

0.46

 convient

0.35

 answer

0.34

 eccl

0.31

 modern

0.31

answer

0.30

nice

0.30

natural

0.30

 astu

0.30

 витра

0.30

Activations Density 0.029%

structured and procedural elements in programming or documentation

answer text

No Comments

No Known Activations