INDEX

Explanations

references to software or coding environments

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

user input separators

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

########.

-0.81

AxisAlignment

-0.56

InitVars

-0.53

ymce

-0.53

 AttributeSet

-0.53

 MainAxisSize

-0.52

reportWebVitals

-0.52

IsMutable

-0.50

ſelves

-0.50

 Conſ

-0.50

POSITIVE LOGITS

 Chwiliwch

0.52

httphttps

0.49

 للاسماء

0.40

qai

0.35

ipedi

0.33

estic

0.33

eqn

0.31

tanleria

0.31

outer

0.31

rips

0.31

Activations Density 0.010%

references to software or coding environments

user input separators

No Comments

No Known Activations