INDEX

Explanations

references to reports and reporting in general

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

report followed by

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

brass

-0.61

DotNetBar

-0.58

XIM

-0.56

infinity

-0.54

swimming

-0.53

 Blackstone

-0.53

ylde

-0.53

switching

-0.53

Nub

-0.52

POSITIVE LOGITS

 report

1.28

 Report

1.15

report

1.14

 REPORT

1.02

Report

1.00

 reports

0.99

 reporte

0.91

REPORT

0.90

 Reports

0.85

 reported

0.77

Activations Density 0.012%

references to reports and reporting in general

report followed by

No Comments

No Known Activations