INDEX

Explanations

terms related to relationships and relational concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Schuster

-0.82

 Villanueva

-0.76

<strong>

-0.75

iama

-0.73

<h1>

-0.73

 Lipp

-0.71

 bianca

-0.70

fizierung

-0.69

lanzen

-0.69

fontSize

-0.68

POSITIVE LOGITS

rel

1.41

rel

1.39

REL

1.16

REL

1.09

Rel

1.08

 Relational

0.98

 lenker

0.90

Rel

0.90

 relational

0.89

 Mako

0.85

Activations Density 0.007%

terms related to relationships and relational concepts

No Comments

No Known Activations