INDEX

Explanations

conjunctions and pronouns, particularly "that" and "which."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.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

ngdoc

-0.47

很容易

-0.43

 varandra

-0.43

dalla

-0.42

 tampoco

-0.42

 négociations

-0.41

 gereken

-0.41

OFFICIAL

-0.41

ారి

-0.41

ستار

-0.40

POSITIVE LOGITS

脚注の使い方

0.87

RegressionTest

0.80

abestanden

0.77

makeText

0.72

 AssemblyProduct

0.71

Geplaatst

0.70

ⓧ

0.69

umab

0.68

enumi

0.68

Vidite

0.68

Activations Density 0.337%

conjunctions and pronouns, particularly "that" and "which."

No Comments

No Known Activations