INDEX

Explanations

terms related to programming definitions and signal handling

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

use followed by comma or number

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 16-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_16/width_65k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.16.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

 schimb

-0.51

rdoba

-0.51

 Lordships

-0.49

czek

-0.46

zioso

-0.46

 compounding

-0.46

vers

-0.45

edly

-0.45

 inconspicuous

-0.44

 oscura

-0.44

POSITIVE LOGITS

 nakalista

0.88

Portály

0.78

حياته

0.75

copg

0.73

ReusableCell

0.71

VersionUID

0.71

Бахар

0.70



0.69

例文帳に追加

0.69

twimg

0.67

Activations Density 0.005%

terms related to programming definitions and signal handling

use followed by comma or number

No Comments

No Known Activations