INDEX

Explanations

programming constructs related to switch cases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

 Monfieur

-0.79

 doubtnut

-0.78

 themſelves

-0.76

 itſelf

-0.74

ibouti

-0.74

 فريبيس

-0.74

 Hadrian

-0.74

 raiſ

-0.73

 poffible

-0.73

 bershka

-0.73

POSITIVE LOGITS

case

2.90

 case

2.40

 Case

2.29

Case

2.23

 CASE

2.14

CASE

2.08

cases

1.85

 cases

1.82

 Cases

1.81

Cases

1.74

Activations Density 0.028%

programming constructs related to switch cases

No Comments

No Known Activations