INDEX

Explanations

references to information availability and its assessment within a context

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

information availability

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

UpInside

-0.41

 térmico

-0.40

 nemlig

-0.40

zwungen

-0.39

 Comprometido

-0.38

 nämlich

-0.35

AppCompat

-0.35

 saites

-0.35

 Grüße

-0.34

">#

-0.34

POSITIVE LOGITS

 information

1.10

information

0.99

 informations

0.97

 Information

0.94

 Data

0.91

 data

0.91

Information

0.84

 INFORMATION

0.82

 info

0.80

informations

0.77

Activations Density 0.096%

references to information availability and its assessment within a context

information availability

No Comments

No Known Activations