INDEX

Explanations

references to experimental methodologies and parameters in scientific research

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

multiple repetitions

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

اریخ

-0.44

Bif

-0.37

Hentet

-0.36

 Infl

-0.35

 Pound

-0.35

 Quid

-0.35

いしい

-0.34

 forked

-0.34

 getClient

-0.33

vorschaubild

-0.33

POSITIVE LOGITS

 kasarigan

0.76

rungsseite

0.58

:✨

0.52

RUnlock

0.48

verwijspagina

0.45

RunWith

0.45

Subviews

0.44

FlatAppearance

0.44

ThroughAttribute

0.42

Variation

0.42

Activations Density 0.165%

references to experimental methodologies and parameters in scientific research

multiple repetitions

No Comments

No Known Activations