INDEX

Explanations

phrases related to programming actions or operations

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

actions and processes

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

mybatisplus

-0.58

Ӕ

-0.47

 Begründung

-0.46

########.

-0.45

andangan

-0.44

SharedDtor

-0.43

 nikdy

-0.42

 nemlig

-0.42

berdayakan

-0.41

 paylaş

-0.41

POSITIVE LOGITS

addPreferredGap

0.56

httphttps

0.47

GEBURTSDATUM

0.43

KURZBESCHREIBUNG

0.41

⋱

0.40

expandindo

0.38

GOING

0.38

otry

0.38

帖最后由

0.37

atting

0.37

Activations Density 0.027%

phrases related to programming actions or operations

actions and processes

No Comments

No Known Activations