INDEX

Explanations

quantifiable terms related to amounts or quantities

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

편

-0.41

GONE

-0.39

>`;

-0.38

 przestr

-0.36

łys

-0.36

 제공

-0.36

 Palis

-0.35

olve

-0.35

 proviene

-0.35

 Cresc

-0.35

POSITIVE LOGITS

IsContent

0.66

 autorytatywna

0.53

 nakalista

0.52

RenderAtEndOf

0.47

UnitTesting

0.45

 AssemblyTitle

0.44

fromnode

0.44

ContentAsync

0.44

曖昧さ回避

0.43

 صوتيه

0.41

Activations Density 0.029%

quantifiable terms related to amounts or quantities

No Comments

No Known Activations