INDEX

Explanations

items related to opportunities, effectiveness, and significance in various contexts

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

explaining outcomes or characteristics

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

DockStyle

-0.38

 geçir

-0.38

PONENTS

-0.35

}`)

-0.34

layın

-0.33

 sfida

-0.33

 murni

-0.31

:])

-0.30

 fought

-0.30

 kämpfen

-0.30

POSITIVE LOGITS

fjspx

0.55

 Monfieur

0.52

reportError

0.50

Diweddarwch

0.47

PYX

0.47

astify

0.46

 ſal

0.46

audiovisuel

0.45

}{@

0.45

 StringTokenizer

0.45

Activations Density 0.131%

items related to opportunities, effectiveness, and significance in various contexts

explaining outcomes or characteristics

No Comments

No Known Activations