INDEX

Explanations

terms related to usage and implementation instructions

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

 Schult

-0.82

er

-0.79

 Hoon

-0.78

DeL

-0.77

']);

-0.76

 Balboa

-0.75

 Delaney

-0.75

 sympy

-0.74

cobra

-0.73

 snows

-0.73

POSITIVE LOGITS

 Usage

1.27

Usage

1.19

 usage

1.18

usage

1.10

usages

1.05

USAGE

1.00

 USAGE

0.98

 usages

0.83

paravant

0.79

싶

0.75

Activations Density 0.040%

terms related to usage and implementation instructions

No Comments

No Known Activations