INDEX

Explanations

phrases that indicate repetition or reference to previous statements

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

reiterating or mentioning past statements

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

 AssemblyTitle

-0.63

 informée

-0.63

 propOrder

-0.61

 surla

-0.57

 vPvB

-0.56

 imagui

-0.56

 otomatig

-0.55

 թվական

-0.54

UnusedPrivate

-0.54

CppCodeGen

-0.54

POSITIVE LOGITS

RectangleBorder

0.56

 reiterated

0.39

 saying

0.37

 reiterate

0.37

GIVEREF

0.37

 stressed

0.32

 λέ

0.31

 reiter

0.30

 stressing

0.30

 federal

0.30

Activations Density 0.028%

phrases that indicate repetition or reference to previous statements

reiterating or mentioning past statements

No Comments

No Known Activations