INDEX

Explanations

forward slashes used in various contexts

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

separators followed by specific words

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

 oreja

-0.42

 nemlig

-0.41

 ervan

-0.40

ツイート

-0.39

 meja

-0.38

 dessus

-0.37

 anzi

-0.37

 conspiracy

-0.36

 ongoing

-0.36

 infatti

-0.36

POSITIVE LOGITS

addCriterion

0.65

 فريبيس

0.65

 Италијани

0.62

MigrationBuilder

0.61

 gynhyrchwyd

0.60

 Autorisations

0.51

NamedQueries

0.51

reportWebVitals

0.51

migrationBuilder

0.50

matchCondition

0.50

Activations Density 0.015%

forward slashes used in various contexts

separators followed by specific words

No Comments

No Known Activations