INDEX

Explanations

structure-related terms and expressions in programming contexts

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

`IIA` or `OBJ`

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

principalTable

-0.66

AccessorTable

-0.63

parsedMessage

-0.62

 صوتيه

-0.60

EDEFAULT

-0.60

 betweenstory

-0.58

:✨

-0.54

numerusform

-0.54

adpleegd

-0.53

principalColumn

-0.53

POSITIVE LOGITS

">//

0.46

 nakalista

0.36

 détail

0.35

 demás

0.34

 مشين

0.33

//

0.33

Spoljašnje

0.32

 barata

0.32

 juos

0.31

上手

0.31

Activations Density 0.402%

structure-related terms and expressions in programming contexts

`IIA` or `OBJ`

No Comments

No Known Activations