INDEX

Explanations

comparative statements about product features and values

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

programming related terms

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

 human

-0.35

 health

-0.32

paroles

-0.31

 Gobierno

-0.31

Dichloropropane

-0.31

inės

-0.30

 Badan

-0.29

 phénomènes

-0.29

 hypothesis

-0.29

DebuggerNonUser

-0.29

POSITIVE LOGITS

ConstraintMaker

0.73

ModelAdmin

0.66

principalColumn

0.62

Espèce

0.56

省市镇

0.56

liceerd

0.56

 NSCoder

0.54

gbaar

0.53

utilisons

0.52

 ExecuteAsync

0.52

Activations Density 0.033%

comparative statements about product features and values

programming related terms

No Comments

No Known Activations