INDEX

Explanations

statistical terminology and references to statistical metrics or functions

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

stat names and values

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

 Ruby

-0.59

PANY

-0.57

 People

-0.56

EEC

-0.56

ruby

-0.55

People

-0.55

bunch

-0.54

 Julie

-0.54

 Lieber

-0.53

POSITIVE LOGITS

 Stat

0.95

Stat

0.93

stat

0.93

 stat

0.92

 Stats

0.85

Stats

0.85

STAT

0.82

 STAT

0.77

 stats

0.77

printStats

0.74

Activations Density 0.019%

statistical terminology and references to statistical metrics or functions

stat names and values

No Comments

No Known Activations