INDEX

Explanations

expressions of recommendation and alternatives

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

alternative options or method

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

erialize

-0.37

########.

-0.34

GF

-0.34

/*

-0.33

 ligiloj

-0.33

Assista

-0.32

PhysRev

-0.31

ждую

-0.31

 semmi

-0.31

 deci

-0.31

POSITIVE LOGITS

 kasarigan

0.68

StoryboardSegue

0.61

delwed

0.58

JMenuBar

0.56

 AssemblyCulture

0.53

 GenerationType

0.53

LLocation

0.52

 BoxDecoration

0.50

setColumns

0.50

WebMethod

0.47

Activations Density 0.084%

expressions of recommendation and alternatives

alternative options or method

No Comments

No Known Activations