INDEX

Explanations

terms related to associations and connections in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_2/width_16k/average_l0_141

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

 Drit

-0.61

jdt

-0.57

dak

-0.57

meld

-0.56

lymp

-0.55

μ

-0.55

ू

-0.53

Bbb

-0.52

Imm

-0.52

Vla

-0.52

POSITIVE LOGITS

 associated

3.02

associated

2.76

Associated

2.57

 Associated

2.53

 ASSOCIATED

2.45

 associate

2.25

 associating

2.19

 associates

2.10

 associ

2.08

 asso

2.05

Activations Density 0.109%

terms related to associations and connections in various contexts

No Comments

No Known Activations