INDEX

Explanations

mentions of research activities or studies being conducted

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

Here, we explain

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

 ſever

-0.51

 myſelf

-0.50

 unknownFields

-0.48

 fevere

-0.46

 textStatus

-0.43

 Blackjack

-0.43

 ſeveral

-0.43

 pleaſure

-0.42

 himſelf

-0.41

 colorWith

-0.40

POSITIVE LOGITS

 Здесь

0.82

 Here

0.81

HERE

0.80

Здесь

0.80

Here

0.79

 here

0.77

 Aquí

0.75

 HERE

0.75

 aquí

0.73

here

0.72

Activations Density 0.045%

mentions of research activities or studies being conducted

Here, we explain

No Comments

No Known Activations