INDEX

Explanations

phrases indicating assertions or claims about truthfulness or reality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_10/width_16k/average_l0_77

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.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

 AutoMapper

-0.81

UserScript

-0.69

amiya

-0.64

ρους

-0.62

 dormitorios

-0.60

TSV

-0.58

bucha

-0.58

RFID

-0.58

OGS

-0.56

Wallis

-0.56

POSITIVE LOGITS

 indeed

0.87

 anzi

0.85

 Bahkan

0.84

 even

0.80

 bahkan

0.80

 sogar

0.79

Bahkan

0.77

indeed

0.74

Indeed

0.72

 hatta

0.72

Activations Density 0.143%

phrases indicating assertions or claims about truthfulness or reality

No Comments

No Known Activations