INDEX

Explanations

assertive statements that indicate conclusions, findings, or observations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 queſta

-0.77

__*/

-0.70

"])

-0.70

PYX

-0.69

 ویکی‌پدی

-0.68

 незавершена

-0.68

parsedMessage

-0.68

<unused74>

-0.66

<unused28>

-0.66

<pad>

-0.66

POSITIVE LOGITS

 that

1.02

 faptul

0.52

 nadzieję

0.51

 bahwa

0.51

 which

0.48

 Tatsache

0.44

 اینکه

0.42

 knows

0.42

 know

0.41

 wiem

0.41

Activations Density 0.831%

assertive statements that indicate conclusions, findings, or observations

No Comments

No Known Activations