INDEX

Explanations

references to measurements or values associated with 's'

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

possessive 's'

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

 myſelf

-0.81

 ſever

-0.78

 pleaſure

-0.78

 ſtate

-0.75

ſelves

-0.74

 purpoſe

-0.74

 ſche

-0.71

 ſtand

-0.70

iſen

-0.69

 ſind

-0.68

POSITIVE LOGITS

0.86

own

0.53

0.51

ğın

0.48

 Spass

0.47

on

0.47

esModule

0.45

 Childs

0.45

and

0.44

Jeografia

0.44

Activations Density 0.013%

references to measurements or values associated with 's'

possessive 's'

No Comments

No Known Activations