INDEX

Explanations

proper nouns, particularly names of people

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

names followed by surnames

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

 betweenstory

-0.73

middels

-0.51

ttemberg

-0.51

 kasarigan

-0.49

吋

-0.48

➯

-0.48

 queſta

-0.46

aneers

-0.45

iterranée

-0.45

↻

-0.45

POSITIVE LOGITS

 nakalista

0.59

Geplaatst

0.54

CEO

0.45

 said

0.42

 spokesperson

0.42

CEO

0.41

 spokesman

0.39

 President

0.39

said

0.38

President

0.38

Activations Density 0.012%

proper nouns, particularly names of people

names followed by surnames

No Comments

No Known Activations