INDEX

Explanations

references to individuals in a document

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

human persons

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

bowl

-0.58

Beat

-0.52

beat

-0.52

 ſta

-0.50

 Reſ

-0.49

 obé

-0.49

alib

-0.49

 Tribut

-0.48

boat

-0.48

Split

-0.48

POSITIVE LOGITS

 persons

1.73

 Persons

1.58

Persons

1.54

 PERSONS

1.40

persons

1.38

 Personen

0.97

 personen

0.87

 perfons

0.85

 personer

0.79

 осіб

0.79

Activations Density 0.006%

references to individuals in a document

human persons

No Comments

No Known Activations

references to individuals in a document

human persons

No Comments

No Known Activations