INDEX

Explanations

comparative phrases indicating magnitude or superiority

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

finding which is greater

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 armband

-0.62

bingen

-0.59

chips

-0.55

 pyjama

-0.54

 ROLL

-0.54

Linux

-0.53

 pyjamas

-0.53

 alibi

-0.52

 timings

-0.52

 sabbat

-0.52

POSITIVE LOGITS

 greater

1.90

greater

1.72

 Greater

1.65

Greater

1.63

GREATER

1.36

 greatest

1.09

greatest

1.05

 groter

0.97

 greate

0.96

 größer

0.94

Activations Density 0.015%

comparative phrases indicating magnitude or superiority

finding which is greater

No Comments

No Known Activations