INDEX

Explanations

phrases that express intentions, particularly those starting with "to."

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

aims to do something

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

󠁴

-0.78

 OMITBAD

-0.70

 estekak

-0.70

 infile

-0.67

 resourceCulture

-0.66

IndentedString

-0.66

 nakalista

-0.65

EDEFAULT

-0.65

locker

-0.65

aconda

-0.64

POSITIVE LOGITS

 aims

0.95

aim

0.90

 aimed

0.82

 bertujuan

0.73

 Tujuan

0.71

 aiming

0.71

 Aims

0.71

 objectif

0.71

 objetivo

0.70

Aim

0.69

Activations Density 0.016%

phrases that express intentions, particularly those starting with "to."

aims to do something

No Comments

No Known Activations