INDEX

Explanations

references to quantitative data or statistics within the text

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-mlp/layer_6/width_131k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.6.hook_mlp_out

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

ῆ

-0.38

Achter

-0.37

mobile

-0.37

LElement

-0.35

 الف

-0.35

vuo

-0.35

 промежу

-0.34

 disappointment

-0.34

ellipsis

-0.34

 turn

-0.34

POSITIVE LOGITS

RegressionTest

0.59

 nakalista

0.57

 AssemblyCulture

0.57

RenderAtEndOf

0.57

IUrlHelper

0.55

expandindo

0.54

QMetaType

0.54

 تكبرها

0.52

FunctionFlags

0.52

 لينك

0.52

Activations Density 0.525%

references to quantitative data or statistics within the text

No Comments

No Known Activations