INDEX

Explanations

terms related to various scientific models and classifications

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

legal and technical 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

 betreft

-0.66

 enfans

-0.62

depends

-0.56

 prêtres

-0.54

 peuples

-0.50

vœ

-0.50

 acompaña

-0.49

ientras

-0.49

ıdır

-0.49

 mergeFrom

-0.49

POSITIVE LOGITS

 applications

0.52

 bezeichneter

0.49

 assignments

0.48

 evaluations

0.46

 presentations

0.43

 descriptions

0.43

 deployments

0.42

 apps

0.42

 solutions

0.41

 elevations

0.41

Activations Density 0.296%

terms related to various scientific models and classifications

legal and technical terms

No Comments

No Known Activations