INDEX

Explanations

terms related to evidence and factual claims

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 ويكيميديا

-0.51

 contextLoads

-0.42

__':

-0.42

 AssemblyVersion

-0.40

saraba

-0.40

SBATCH

-0.40

 '-')

-0.40

lusconi

-0.38

cycl

-0.38

trische

-0.37

POSITIVE LOGITS

各地

0.58

 nationwide

0.58

 distributors

0.52

 aanges

0.47

@[+][

0.47

 municipalities

0.47

加盟

0.47

 distributed

0.46

 delegates

0.45

 Distributors

0.44

Activations Density 0.705%

terms related to evidence and factual claims

No Comments

No Known Activations