INDEX

Explanations

identifiers and types used in programming code

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

id; thread; socket; Ac

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

AndEndTag

-0.76

 pouvoit

-0.63

 ſever

-0.57

+#+#

-0.56

ſelves

-0.54

NameInMap

-0.52

LookAnd

-0.52

 auroit

-0.52

RenderAtEndOf

-0.50

 Italijanski

-0.50

POSITIVE LOGITS

UniformLocation

0.41

 Mach

0.34

bogen

0.34

 Kessel

0.33

ko

0.33

Gil

0.33

Sal

0.32

logic

0.32

Ộ

0.32

hoeddwyd

0.32

Activations Density 0.115%

identifiers and types used in programming code

id; thread; socket; Ac

No Comments

No Known Activations