INDEX

Explanations

words related to a negative or problematic sense of integrity or honesty

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

starting with "dis," "imp," or "squ"

np_acts-logits-general · gemini-2.0-flash

dis, imp, squ, uncons, mis

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 Geduld

-0.52

Today

-0.52

Архівовано

-0.51

parsedMessage

-0.51

>");

-0.50

 Erbe

-0.50

EndInit

-0.49

UNIX

-0.48

########.

-0.48

 Today

-0.46

POSITIVE LOGITS

dis

0.50

pre

0.50

Squ

0.49

squ

0.48

uns

0.48

 smart

0.48

hamdu

0.48

 hyper

0.47

 fire

0.47

 bright

0.47

Activations Density 0.613%

words related to a negative or problematic sense of integrity or honesty

starting with "dis," "imp," or "squ"

dis, imp, squ, uncons, mis

No Comments

No Known Activations