INDEX

Explanations

the word "what" in various contexts

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

no matter what

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

 ModelExpression

-0.54

Personendaten

-0.53

berdayakan

-0.52

FieldBuilder

-0.51

ScopeManager

-0.50

 Comprometido

-0.49

enderror

-0.47

Descripció

-0.47

ніципалі

-0.47

 ostavi

-0.47

POSITIVE LOGITS

the

0.58

 whatever

0.55

any

0.51

 Manning

0.47

 qualunque

0.47

Any

0.47

 Whatever

0.46

 disobey

0.46

Whatever

0.46

whatever

0.46

Activations Density 0.005%

the word "what" in various contexts

no matter what

No Comments

No Known Activations