INDEX

Explanations

interrogative phrases, particularly those beginning with "how" and "why."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 14-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_14/width_16k/average_l0_84

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.14.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

 occaf

-0.88

 greateſt

-0.87

 itſelf

-0.87

 preſent

-0.84

 raft

-0.82

 poffe

-0.82

 ftate

-0.82

 Majefty

-0.79

 purpoſe

-0.79

 Gegenteil

-0.79

POSITIVE LOGITS

how

1.61

How

1.43

HOW

1.38

How

1.28

HOW

1.19

how

1.01

 much

0.92

 Howe

0.90

 Howell

0.88

 what

0.87

Activations Density 0.070%

interrogative phrases, particularly those beginning with "how" and "why."

No Comments

No Known Activations

interrogative phrases, particularly those beginning with "how" and "why."

No Comments

No Known Activations