INDEX

Explanations

code-related terms and function calls

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

GEBURTSDATUM

-0.60

ThroughAttribute

-0.52

MessageTagHelper

-0.52

PyExc

-0.48

reportWebVitals

-0.48

VolleyError

-0.46

outSlope

-0.46

 AssemblyTitle

-0.46

 pinulongan

-0.45

'{@

-0.45

POSITIVE LOGITS

yntaxException

0.57

 diseñar

0.48

 gyhoeddwyd

0.48

 otomatig

0.46

 planification

0.42

 diseños

0.41

 diseñado

0.41

 diseño

0.40

PerformLayout

0.40

 concevoir

0.40

Activations Density 0.052%

code-related terms and function calls

No Comments

No Known Activations