INDEX

Explanations

references to arguments in code or programming contexts

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

programming arguments

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

 Excited

-0.35

 empty

-0.35

 pride

-0.34

 hotspot

-0.33

 excited

-0.33

 stabilize

-0.33

Praying

-0.32

Tento

-0.32

ufac

-0.32

 plan

-0.31

POSITIVE LOGITS

arguments

1.59

 arguments

1.26

 argument

1.13

argument

1.03

 argumento

0.97

 Arguments

0.96

Arguments

0.90

 Argument

0.87

 argumentos

0.85

 arguing

0.84

Activations Density 0.002%

references to arguments in code or programming contexts

programming arguments

No Comments

No Known Activations