INDEX

Explanations

proper names of individuals

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

first names followed by last names

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

principalTable

-0.49

 nakalista

-0.47

kháu

-0.47

ThroughAttribute

-0.44

 مشين

-0.44

 Roskov

-0.43

 Infórmanos

-0.42

isContained

-0.41

 estekak

-0.40

RTLD

-0.40

POSITIVE LOGITS

iſen

0.45

 Audiodateien

0.44

 CreateTagHelper

0.43

AddTagHelper

0.41

 houſe

0.41

dingen

0.40

Biografia

0.39

 pleaſure

0.39

 vence

0.39

createClass

0.38

Activations Density 0.028%

proper names of individuals

first names followed by last names

No Comments

No Known Activations