INDEX

Explanations

conversational elements and expressions of uncertainty

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

"i" followed by observation or uncertainty

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

 aast

-0.36

råd

-0.32

 kuah

-0.32

いなかった

-0.32

 Viitattu

-0.31

历史上

-0.29

|+\

-0.29

 ketahui

-0.29

 elevados

-0.29

ніципа

-0.29

POSITIVE LOGITS

parsedMessage

0.65

 témoig

0.65

MessageOf

0.60

 ControllerBase

0.60

MessageTagHelper

0.59

GOTREF

0.59

 queſta

0.58

 المعيارى

0.58

 Symptom

0.57

ésultats

0.57

Activations Density 0.057%

conversational elements and expressions of uncertainty

"i" followed by observation or uncertainty

No Comments

No Known Activations