INDEX

Explanations

references to acts of obtaining or seeking information or resources

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

discover policy seems reluctant

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

 témoig

-0.63

 <<<<<<<<<<<<<<

-0.62

 Geſch

-0.60

iſche

-0.59

ագրություններ

-0.59

 deſſen

-0.58

 dieſes

-0.57

linawan

-0.57

+#+#

-0.57

 InputDecoration

-0.56

POSITIVE LOGITS

Pu

0.25

竟然

0.25

UpdateCommand

0.25

se

0.24

͘

0.24

 darn

0.24

 pull

0.23

Pi

0.23

EV

0.23

Pi

0.23

Activations Density 0.048%

references to acts of obtaining or seeking information or resources

discover policy seems reluctant

No Comments

No Known Activations