INDEX

Explanations

terms related to irregularities and inconsistencies in behavior or performance

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

inconsistent and irregular

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

 esternos

-0.48

tiamo

-0.47

InSection

-0.44

GEBURTS

-0.44

Reach

-0.43

Wohn

-0.42

Nose

-0.42

 defStyleAttr

-0.41

odenal

-0.41

enheim

-0.41

POSITIVE LOGITS

 erratic

0.72

 fluctuating

0.67

 unpredictable

0.63

 unreliable

0.63

 unstable

0.61

 irregular

0.58

 sporadic

0.57

 sporad

0.56

 uncertain

0.56

 uneven

0.56

Activations Density 0.054%

terms related to irregularities and inconsistencies in behavior or performance

inconsistent and irregular

No Comments

No Known Activations