INDEX

Explanations

dialogue and quotations in the text

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

comma followed by a word

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

 Organ

-0.35

-0.33

-0.32

 Modern

-0.31

'/

-0.30

 {});

-0.29

 Fund

-0.29

POSITIVE LOGITS

 nakalista

0.78

AndEndTag

0.73

 disambiguazione

0.67

WriteTagHelper

0.66

ðsíða

0.63

 sambil

0.61

MemoryWarning

0.59

AddTagHelper

0.59

 nhắc

0.59

нгред

0.57

Activations Density 0.007%

dialogue and quotations in the text

comma followed by a word

No Comments

No Known Activations