INDEX

Explanations

references to smells and sensory experiences associated with various contexts

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

sense of smell

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

 Darcy

-0.59

 carbonic

-0.56

 Mackey

-0.53

づけ

-0.52

 Pacers

-0.51

 prior

-0.51

背影

-0.50

Monarch

-0.50

bondale

-0.50

ACI

-0.50

POSITIVE LOGITS

smell

0.94

Smell

0.90

 smell

0.89

 smells

0.84

 Smell

0.84

 smelling

0.76

 smelled

0.69

sme

0.64

 smelly

0.64

Sme

0.61

Activations Density 0.003%

references to smells and sensory experiences associated with various contexts

sense of smell

No Comments

No Known Activations