INDEX

Explanations

discussions about evidence and claims in legal contexts

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

seeking specific details

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

 surla

-0.73



-0.62

AlterField

-0.58

'{@

-0.55

Rüyada

-0.54

MigrationBuilder

-0.50

CppMethod

-0.48

ьажоргаш

-0.46

 BoxDecoration

-0.46

 disambiguazione

-0.45

POSITIVE LOGITS

 konkrete

0.41

 confronti

0.40

 specifics

0.39

 concreto

0.38

 aislada

0.36

 Ausnahme

0.36

 конкре

0.35

 Infór

0.35

 cụ

0.35

 concrete

0.35

Activations Density 0.182%

discussions about evidence and claims in legal contexts

seeking specific details

No Comments

No Known Activations