INDEX

Explanations

references to in-person interactions and processes

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

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_32k/average_l0_76

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

float32

Hook Name

blocks.12.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

 réglable

-0.54

 Vikipedi

-0.54

jestel

-0.48

gyz

-0.48

UserScript

-0.48

 particulières

-0.46

 silicona

-0.46

 trattato

-0.46

StructEnd

-0.45

isNameExpr

-0.45

POSITIVE LOGITS

 physical

1.18

 physically

1.08

physical

1.00

PHYSICAL

0.96

 PHYSICAL

0.96

Physical

0.96

 Physical

0.95

 Physically

0.92

 presencial

0.91

 fysis

0.86

Activations Density 0.208%

references to in-person interactions and processes

No Comments

No Known Activations