INDEX

Explanations

phrases indicating additional information or contrasts

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

adding further points

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 pouvoit

-0.41

 miſ

-0.39

dessous

-0.38

 nebu

-0.36

 Piotr

-0.35

 deſt

-0.35

الصفحه

-0.35

ffions

-0.35

 själv

-0.35

 standing

-0.34

POSITIVE LOGITS

何より

0.58

abestanden

0.51

 للاسماء

0.49

 Paglinawan

0.48

MessageOf

0.47

 CreateTagHelper

0.46

RTDA

0.46

Hochspringen

0.46

complexContent

0.45

dafx

0.45

Activations Density 0.441%

phrases indicating additional information or contrasts

adding further points

No Comments

No Known Activations