INDEX

Explanations

references to sections and theorems within a mathematical context

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

following numbered list items

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

featureID

-0.54

tagHelperRunner

-0.41

-0.40

цездатний

-0.40

 Chwiliwch

-0.39

ỡng

-0.36

 gangguan

-0.36

 мәкал

-0.35

 hardening

-0.35

 forefathers

-0.35

POSITIVE LOGITS

 GenerationType

0.53

tableFuture

0.49

oad

0.47

bkz

0.45

localctx

0.45

PostInfinity

0.44

transQ

0.43

 cref

0.42

yntaxException

0.42

 منز

0.41

Activations Density 0.040%

references to sections and theorems within a mathematical context

following numbered list items

No Comments

No Known Activations