INDEX

Explanations

mentions of characters and entities in a narrative context

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

robots and bots

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

Rüyada

-0.41

 كومونز

-0.40

 Италијани

-0.39

ViewFeatures

-0.39

 AssemblyCulture

-0.37

省市镇

-0.36

windowFixed

-0.35

AnchorStyles

-0.35

película

-0.35

 насељу

-0.34

POSITIVE LOGITS

 robots

0.73

 robot

0.71

 chatbot

0.65

 Robots

0.64

 robotic

0.61

haikusbot

0.61

 Robot

0.59

 robo

0.58

robo

0.58

 ChatGPT

0.57

Activations Density 0.176%

mentions of characters and entities in a narrative context

robots and bots

No Comments

No Known Activations