INDEX

Explanations

programming-related terminology and concepts, especially those pertaining to functions and control structures

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

technical code structures

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

 blijkt

-0.42

 arşivlendi

-0.41

 leeftijd

-0.39

principalColumn

-0.36

 strijd

-0.35

Rgds

-0.35

 płyn

-0.35

 géant

-0.35

 Accesat

-0.34

 voorbeeld

-0.33

POSITIVE LOGITS

UrlResolution

0.59

CppMethod

0.58

 nonUne

0.52

httphttps

0.51

IUrlHelper

0.49

 pinulongan

0.48

âteau

0.45

toISOString

0.45

ukkan

0.43

InlineData

0.42

Activations Density 0.020%

programming-related terminology and concepts, especially those pertaining to functions and control structures

technical code structures

No Comments

No Known Activations