INDEX

Explanations

terms related to the characterization of subjects or processes

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

"Character" followed by suffixes like "ization"

np_acts-logits-general · gemini-2.0-flash

character and words like characterization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

GOING

-0.70

IBarButtonItem

-0.70

numerusform

-0.66

borderSide

-0.66

rivial

-0.65

 citoy

-0.65

 Aldo

-0.64

 Odyssey

-0.64

 tutup

-0.64

chengladbach

-0.63

POSITIVE LOGITS

 characters

0.71

ization

0.69

 CHAR

0.66

chartInstance

0.66

 Characters

0.65

 character

0.64

 CHARACTER

0.62

achter

0.61

 Char

0.59

Characters

0.58

Activations Density 0.154%

terms related to the characterization of subjects or processes

"Character" followed by suffixes like "ization"

character and words like characterization

No Comments

No Known Activations