INDEX

Explanations

references to media publications and platforms

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

media websites or news outlets

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

 anaknya

-0.42

 смотря

-0.39

combinations

-0.39

 boneka

-0.38

blind

-0.38

negate

-0.38

far

-0.38

 väh

-0.38

WithIOException

-0.38

 EPISODE

-0.37

POSITIVE LOGITS

Tembelea

0.84



0.61

 незавершена

0.59

 estekak

0.57

tagHelperRunner

0.57

:+:

0.57

AndroidJUnit

0.55

+#+#

0.51

 resourceCulture

0.50

StoryboardSegue

0.50

Activations Density 0.015%

references to media publications and platforms

media websites or news outlets

No Comments

No Known Activations