INDEX

Explanations

references to historical or past uses of locations and structures

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

railroad, canal, or railway corridors

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

ResponseWriter

-0.46

setVerticalGroup

-0.42

响起

-0.42

 cerchi

-0.41

 Embaj

-0.41

Personendaten

-0.41

 embaj

-0.40

 bahía

-0.40

RegressionTest

-0.40

 Gestión

-0.39

POSITIVE LOGITS

 سكانية

0.49

 Biôgrafia

0.46

最快更新

0.45

careous

0.44

хьтан

0.41

pogon

0.41

 HasFactory

0.39

IUrlHelper

0.39

 '-':

0.39

principalTable

0.39

Activations Density 0.021%

references to historical or past uses of locations and structures

railroad, canal, or railway corridors

No Comments

No Known Activations