INDEX

Explanations

key terms related to improvement and success in various contexts

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

abstract concepts and properties

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

MLLoader

-0.49

Хьажоргаш

-0.45

UnusedPrivate

-0.45

 Administrativna

-0.43

 NSCoder

-0.42

>{@

-0.40

sizeCache

-0.40

createCanvas

-0.39

 tetapi

-0.38

ricians

-0.38

POSITIVE LOGITS

GOTREF

0.44

 snippetHide

0.44

.”

0.43

 تضيفلها

0.42

Cordialement

0.41

قایناق‌لار

0.41

पया

0.40

ResumeLayout

0.40

<eos>

0.40

ември

0.39

Activations Density 0.027%

key terms related to improvement and success in various contexts

abstract concepts and properties

No Comments

No Known Activations