INDEX

Explanations

references to key actions and instructions in a procedural or narrative context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.12.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

 kasarigan

-0.56

CloseOperation

-0.51

LookAnd

-0.49

 esternos

-0.49

MouseOver

-0.49

 AppCompatTheme

-0.49

 intptr

-0.49

CppMethod

-0.48

 claim

-0.48

riller

-0.48

POSITIVE LOGITS

cektir

0.58

 GenerationType

0.57

 wireType

0.55

InputTagHelper

0.53

MigrationBuilder

0.52

 itſelf

0.52

 Esau

0.51

 Houſe

0.50

berd

0.49

 saites

0.49

Activations Density 0.205%

references to key actions and instructions in a procedural or narrative context

No Comments

No Known Activations