INDEX

Explanations

elements related to formatted lists or itemized information

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

special tokens or formatting

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

SequentialGroup

-0.36

 meal

-0.35

hyd

-0.32

URERS

-0.32

 relief

-0.31

 معنا

-0.30

ņi

-0.30

odo

-0.29

 सी

-0.29

เลย

-0.29

POSITIVE LOGITS

setVerticalGroup

0.74

MessageTagHelper

0.71

 nahilalakip

0.70

ViewFeatures

0.70

AnchorTagHelper

0.68

 AssemblyCompany

0.67

 propOrder

0.66

abestanden

0.65

 &___

0.64

Personensuche

0.62

Activations Density 0.117%

elements related to formatted lists or itemized information

special tokens or formatting

No Comments

No Known Activations