INDEX

Explanations

relationships and dependencies between various concepts or aspects

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

for obtaining, for assessing

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

 <>",

-0.67

 protoimpl

-0.67

 հղումներ

-0.66

+#+#

-0.66

uxxxx

-0.66

tonode

-0.66

Життєпис

-0.64

Билгалдахарш

-0.63

⟬

-0.63

windowFixed

-0.62

POSITIVE LOGITS

↑

0.33

 activities

0.29

 przyczyn

0.29

 Bestandteil

0.28

 prevent

0.28

 this

0.28

誉

0.28

 Kapitel

0.27

 activity

0.27

 Unterneh

0.26

Activations Density 0.079%

relationships and dependencies between various concepts or aspects

for obtaining, for assessing

No Comments

No Known Activations