INDEX

Explanations

terms related to instructions or guidelines

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_11/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

MessageTagHelper

-0.45

 poffible

-0.37

 ब्रेकडाउन

-0.35

 Pink

-0.34

 AssemblyTitle

-0.34

 intptr

-0.34

</thead>

-0.33

 invokingState

-0.33

ſp

-0.33

 gars

-0.32

POSITIVE LOGITS

extField

0.59

delwed

0.57

➌

0.54

 étoit

0.50

HostException

0.49

 endblock

0.46

 Füße

0.46

 mwenye

0.46

endpush

0.45

 wikipagina

0.44

Activations Density 0.592%

terms related to instructions or guidelines

No Comments

No Known Activations