INDEX

Explanations

references to methodologies or strategies in discussion

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 soldati

-0.77

 Hickey

-0.74

tys

-0.69

 Munro

-0.67

 admiten

-0.67

enumerate

-0.66

 LeBlanc

-0.66

 Crum

-0.65

 arson

-0.65

Mur

-0.65

POSITIVE LOGITS

 approaches

2.26

 approach

2.04

 Approach

2.01

 APPROACH

2.01

 Approaches

1.99

Approach

1.97

approach

1.83

 approached

1.71

Appro

1.55

 approaching

1.50

Activations Density 0.047%

references to methodologies or strategies in discussion

No Comments

No Known Activations