INDEX

Explanations

negative assertions related to legal rulings and motions

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

not doing or finding

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

Negative Logits

stylers

-0.46

出版年

-0.43

 ""],

-0.38

führer

-0.37

開

-0.37

jura

-0.36

开

-0.36

★★★★☆

-0.36

koper

-0.36

 getItemId

-0.36

POSITIVE LOGITS

帖最后由

0.56

AutoField

0.54

 صوتيه

0.52

 cherchés

0.50

 averted

0.48

الحياه

0.47

VYMaps

0.47

الدراسه

0.45

хівовано

0.44

Демографія

0.43

Activations Density 0.435%

negative assertions related to legal rulings and motions

not doing or finding

No Known Activations