INDEX

Explanations

terms related to coding and programming challenges

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.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

BeginContext

-0.68

ViewFeatures

-0.66

elemField

-0.65

/*

-0.65

новништво

-0.61

 linkovi

-0.58

 noDo

-0.58

 >=",

-0.58

همچنین

-0.56

 dabei

-0.53

POSITIVE LOGITS

 alternative

1.23

 alternatives

1.09

 Alternative

1.06

alternative

1.05

 ALTERNATIVE

1.04

Alternative

1.01

 alternativa

1.01

 workaround

1.00

 resorted

1.00

 resorting

0.99

Activations Density 0.612%

terms related to coding and programming challenges

No Comments

No Known Activations