INDEX

Explanations

phrases indicating instructions or requirements

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

This is or means

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

 виправивши

-0.65

 Мексичка

-0.48

 voire

-0.44

Datuak

-0.44

uidado

-0.43

حوالہ

-0.39

Хьажоргаш

-0.39

 nabíz

-0.39

Hochspringen

-0.39

 Italijanski

-0.39

POSITIVE LOGITS

 whoſe

0.52

colgroup

0.47

 tranſ

0.45

atimes

0.45

RTGC

0.45

 disambiguazione

0.44

clearTimeout

0.43

 AppBar

0.42

Diana

0.42

 ſhe

0.42

Activations Density 0.030%

phrases indicating instructions or requirements

This is or means

No Comments

No Known Activations