INDEX

Explanations

terms related to the concept of orientation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 17-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_17/width_65k/average_l0_125

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.17.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

 autorytatywna

-0.66

ylon

-0.65

 HasFactory

-0.64

 NSCoder

-0.64

falls

-0.62

mybatisplus

-0.62

ldquo

-0.61

<eos>

-0.60

 uParam

-0.59

hud

-0.58

POSITIVE LOGITS

 orientations

1.32

 ORIENT

1.23

 Orient

1.18

orientation

1.09

 orientation

1.09

ORIENT

1.09

 Orientation

1.05

 orient

1.04

Orient

1.02

 Oriented

1.02

Activations Density 0.021%

terms related to the concept of orientation

No Comments

No Known Activations