INDEX

Explanations

frequent interactions or references within a dialogue context

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

user parenthesis

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 الرياضيه

-0.78

Jeografia

-0.73

 שוליים

-0.68

StoryboardSegue

-0.60

Hentet

-0.60

ніципалі

-0.60

OGND

-0.57

 ivelany

-0.57

canestro

-0.56

SharedCtor

-0.55

POSITIVE LOGITS

mbra

0.34

0.33

OFF

0.33

dania

0.32

[*]

0.32

INTO

0.32

 gator

0.32

afir

0.31

ียม

0.31

Deja

0.30

Activations Density 0.005%

frequent interactions or references within a dialogue context

user parenthesis

No Comments

No Known Activations