INDEX

Explanations

legal terms and references in context

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

event outcomes and sequences

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

 pinulongan

-0.56

 snippetHide

-0.51

kadot

-0.49

XmlAccessType

-0.48

Дереккөздер

-0.47

getDay

-0.47

SpringBootTest

-0.46

Paj

-0.45

ErrUnexpected

-0.45

vj

-0.45

POSITIVE LOGITS

AutoField

0.47

 AssemblyTitle

0.47

RenderAtEndOf

0.41

ParallelGroup

0.38

 Verpflichtung

0.38

 człowie

0.36

 فريبيس

0.35

 wyna

0.35

 zło

0.34

 Decke

0.34

Activations Density 0.058%

legal terms and references in context

event outcomes and sequences

No Comments

No Known Activations