INDEX

Explanations

phrases and words relating to comparisons or metaphors

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

single words

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

ingeki

-0.48

principalColumn

-0.47

Története

-0.45

Jeografia

-0.44

ULTS

-0.44

спользование

-0.44

 CallOverrides

-0.44

expandindo

-0.43

dali

-0.43

partiet

-0.43

POSITIVE LOGITS

 defaultstate

0.52

HttpPut

0.41

0.39

 mold

0.39

 BoxDecoration

0.38

 BoxFit

0.38

RRect

0.38

Datuak

0.37

Parcelize

0.36

AddTagHelper

0.35

Activations Density 0.035%

phrases and words relating to comparisons or metaphors

single words

No Comments

No Known Activations