INDEX

Explanations

terms related to news broadcasting and production processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 37-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_37/width_16k/average_l0_124

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.37.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

#+#

-0.47

ngrx

-0.47

 ExecuteAsync

-0.46

췄

-0.43

 Taktlose

-0.42

complexContent

-0.40

AndroidJUnit

-0.40

AxisAlignment

-0.40

httphttps

-0.39

providedIn

-0.39

POSITIVE LOGITS

 usually

3.63

 often

3.22

usually

3.14

 typically

3.09

 Usually

3.08

Usually

3.06

 обычно

2.83

 geralmente

2.81

often

2.80

 biasanya

2.78

Activations Density 3.677%

terms related to news broadcasting and production processes

No Comments

No Known Activations