INDEX

Explanations

coding-related terms, particularly those associated with ESLint and build processes

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

 CreateTagHelper

-0.65

OOTDTY

-0.55

 ligiloj

-0.51

المكان

-0.48

ConstraintMaker

-0.47

 EconPapers

-0.46

getDoctrine

-0.46

InstrumentedTest

-0.44

MessageState

-0.44

 dazu

-0.42

POSITIVE LOGITS

IndentedString

0.52

globals

0.51

VIATIONS

0.50

 AssemblyCompany

0.48

 globals

0.47

jshint

0.46

import

0.45

 TextAlign

0.45

outSlope

0.44

 AssemblyTitle

0.44

Activations Density 0.002%

coding-related terms, particularly those associated with ESLint and build processes

No Comments

No Known Activations