INDEX

Explanations

references to internal and external structures or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

 McClure

-0.88

displayquote

-0.82

enumi

-0.80

 coupler

-0.76

 équipée

-0.75

 Certi

-0.74

 inéd

-0.74

 spéciaux

-0.73

kunjung

-0.73

 Moab

-0.73

POSITIVE LOGITS

 Internal

2.26

 internal

2.25

Internal

2.12

 INTERNAL

1.99

internal

1.93

 internally

1.84

INTERNAL

1.75

 internals

1.52

 internas

1.50

 interno

1.47

Activations Density 0.047%

references to internal and external structures or concepts

No Comments

No Known Activations