INDEX

Explanations

references to specific cultural or pop culture elements and their contexts

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

specific phrases

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

sizeCache

-0.53

 resourceCulture

-0.52

DebuggerNonUser

-0.51

 Roskov

-0.47

tvguidetime

-0.47

lüğ

-0.46

Tikang

-0.46

webElementXpaths

-0.45

GOTREF

-0.45

رشف

-0.44

POSITIVE LOGITS

 Réponses

0.52

 Confessions

0.45

 ""],

0.44

 simplifié

0.43

mbic

0.43

 στιγ

0.41

 ujednoznacz

0.41

 CONNECT

0.40

+#+#

0.40

 anch

0.37

Activations Density 0.109%

references to specific cultural or pop culture elements and their contexts

specific phrases

No Comments

No Known Activations