INDEX

Explanations

instances of humor and communication styles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_11/width_16k/average_l0_80

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

Naissance

-0.58

includegraphics

-0.55

RRect

-0.52

plona

-0.52

 fotogra

-0.51

 recens

-0.50

WaitGroup

-0.49

SPEC

-0.49

matsu

-0.48

 Дата

-0.48

POSITIVE LOGITS

 uttered

1.08

 spoken

1.00

 words

0.95

 utterances

0.88

 uttering

0.88

 speech

0.88

 talking

0.81

 Spoken

0.81

 worded

0.80

peech

0.79

Activations Density 0.521%

instances of humor and communication styles

No Comments

No Known Activations