INDEX

Explanations

phrases that suggest exploration, examination, or analysis of concepts

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

Let us

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

 queſta

-0.81

MLLoader

-0.77

 purpoſe

-0.71

 houſe

-0.69

 Houſe

-0.68

 Reſ

-0.66

 ſta

-0.66

 ſte

-0.64

 Majefty

-0.64

fjspx

-0.63

POSITIVE LOGITS

 ourselves

0.60

 here

0.47

Vedi

0.47

Vamos

0.44

Давайте

0.44

 some

0.43

 Vamos

0.42

 this

0.42

Let

0.42

смотрим

0.41

Activations Density 0.267%

phrases that suggest exploration, examination, or analysis of concepts

Let us

No Comments

No Known Activations