INDEX

Explanations

references to case studies or cases

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

testens

-0.56

redients

-0.54

 Whittaker

-0.54

 dirigir

-0.52

strator

-0.52

 Gibbons

-0.52

 hydrodynamic

-0.52

transparent

-0.52

SDP

-0.52

ambiguation

-0.51

POSITIVE LOGITS

Case

1.23

 case

1.19

 Case

1.18

case

1.18

 CASE

1.05

CASE

1.05

 cases

0.98

 Cases

0.91

cases

0.87

Cases

0.87

Activations Density 0.070%

references to case studies or cases

No Comments

No Known Activations