INDEX

Explanations

statements, claims, and reported speech indicating communication or assertion

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

reporting verbs followed by conjunctions

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

 stary

-0.35

UrlResolution

-0.34

 Rota

-0.34

 partita

-0.33

 denial

-0.32

tals

-0.32

 Mane

-0.32

 Bare

-0.32

 rural

-0.32

 Liar

-0.32

POSITIVE LOGITS

MigrationBuilder

0.69

astéro

0.61

 indígen

0.59

 onPostExecute

0.59

ViewImports

0.58

 increí

0.57

 виправивши

0.55

Décès

0.53

 CanadaChoose

0.53

 noDo

0.53

Activations Density 0.065%

statements, claims, and reported speech indicating communication or assertion

reporting verbs followed by conjunctions

No Comments

No Known Activations