INDEX

Explanations

occurrences of utility-related functions and structures in code

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

start of turn user

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

 pleaſure

-0.66

 surla

-0.63

 ddelweddau

-0.62

 houſe

-0.61

 ſta

-0.60

ſelf

-0.60

AddTagHelper

-0.58

-------

-0.56

 deſt

-0.56

 queſta

-0.56

POSITIVE LOGITS

 programação

0.44

 emprender

0.43

 sorpresas

0.40

 inspiração

0.40

 sapere

0.39

 विश्वसनीयता

0.38

 SCOPE

0.38

 noiva

0.38

 madrugada

0.37

 imaginación

0.37

Activations Density 0.001%

occurrences of utility-related functions and structures in code

start of turn user

No Comments

No Known Activations