INDEX

Explanations

references to software or system configurations

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

programming module exports

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

flashdata

-0.45

 становника

-0.38

ueroa

-0.35

 חש

-0.35

fromnode

-0.35

libft

-0.34

chwyd

-0.34

*/].

-0.34

 betre

-0.34

 Observation

-0.33

POSITIVE LOGITS

NgModule

0.60

 snippetHide

0.59

DockStyle

0.58

 kasarigan

0.57

Източници

0.53

Бележки

0.49

 EconPapers

0.48

contentLoaded

0.48

AccessorTable

0.48

WireFormatLite

0.47

Activations Density 0.000%

references to software or system configurations

programming module exports

No Comments

No Known Activations