INDEX

Explanations

terms related to biological processes or interactions involving proteins

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.46

mede

-0.42

 Hache

-0.41

 calendriers

-0.40

menuStrip

-0.39

 cầm

-0.39

 vielmehr

-0.38

ervlak

-0.36

 geworden

-0.35

 Carey

-0.35

POSITIVE LOGITS

 Италијани

0.71

GTCX

0.67

GOTREF

0.64

Rohy

0.57

BufferException

0.57

 smtplib

0.56

 snippetHide

0.55

 الرياضيه

0.55

 AssemblyProduct

0.51

 henvisninger

0.51

Activations Density 0.156%

terms related to biological processes or interactions involving proteins

No Comments

No Known Activations