INDEX

Explanations

negative values or indicators

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

section identifiers

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 pleaſure

-0.53

 faſt

-0.52

 himſelf

-0.52

 leſs

-0.44

 themſelves

-0.43

 psicológico

-0.42

 itſelf

-0.41

ſelf

-0.41

 ſeveral

-0.39

 AssemblyCompany

-0.38

POSITIVE LOGITS

PageIndex

0.65

0.56

oa̍t

0.55

 CanadaChoose

0.52

StructEnd

0.52

RegistryLite

0.52

 pinulongan

0.51

nloa

0.50

 Normdatei

0.50

的特

0.50

Activations Density 0.061%

negative values or indicators

section identifiers

No Comments

No Known Activations