INDEX

Explanations

say "the"

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_32.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.32.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 procedures

-0.26

 papers

-0.26

 soluble

-0.25

_Ptr

-0.25

 fing

-0.25

âĬĩ

-0.25

lew

-0.24

 gesture

-0.24

çļĦåĬ¨ä½ľ

-0.24

ç¤¾ä¼ļç§ĳåŃ¦

-0.24

POSITIVE LOGITS

ç¼ĺ

0.35

Don

0.27

åĪ°åºķ

0.27

ins

0.26

Don

0.26

è¯¤

0.26

åİ¥

0.25

DON

0.25

adata

0.25

ç·£

0.24

Activations Density 0.001%

say "the"

No Known Activations