INDEX

Explanations

API authentication tokens and related parameters

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

user input sections

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

Pim

-0.51

MOT

-0.50

MIR

-0.50

Output

-0.49

MOT

-0.49

cora

-0.48

 зато

-0.48

dorp

-0.48

 OUTPUT

-0.47

haviours

-0.47

POSITIVE LOGITS

 informée

0.43

 Kenntnis

0.41

 Inscrivez

0.38

 henvis

0.38

 atasan

0.37

RefNanny

0.37

 artificiales

0.36

berdayakan

0.36

jsonwebtoken

0.36

 vervangen

0.36

Activations Density 0.004%

API authentication tokens and related parameters

user input sections

No Comments

No Known Activations