INDEX

Explanations

references to variable names or identifiers in programming contexts

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

key existence

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_32/width_131k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.32.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

 snippetHide

-0.79

AndEndTag

-0.75

ValueStyle

-0.75

脚注の使い方

-0.74

nloa

-0.73

 CreateTagHelper

-0.71



-0.70

TagMode

-0.68

RegressionTest

-0.66

isContained

-0.66

POSITIVE LOGITS

 keys

0.43

 clés

0.43

key

0.40

 exist

0.39

 existing

0.37

 exists

0.37

 existed

0.36

 terdapat

0.36

 appeared

0.35

 respectively

0.34

Activations Density 0.007%

references to variable names or identifiers in programming contexts

key existence

No Comments

No Known Activations