INDEX

Explanations

expressions of apology or regret

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

expressing apologies

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

NSM

-0.54

calfe

-0.53

 MutableLiveData

-0.49

MenuInflater

-0.48

 egli

-0.47

 iscri

-0.46

modelBuilder

-0.46

 maș

-0.46

DCE

-0.46

BCA

-0.46

POSITIVE LOGITS

 SORRY

1.09

 sorry

0.98

sorry

0.97

 Sorry

0.96

Sorry

0.94

 désolés

0.63

 Désolé

0.58

 apologised

0.57

 sorri

0.56

 noDo

0.55

Activations Density 0.003%

expressions of apology or regret

expressing apologies

No Comments

No Known Activations