INDEX

Explanations

symbols, formatting, or structure related to document citations and references

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

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

MigrationBuilder

-0.81

OGND

-0.71

 TextInputType

-0.69

tanleria

-0.59

 AssemblyCompany

-0.58

RTHOOK

-0.56

ToBounds

-0.55

 emailAlready

-0.53

transQ

-0.52

󠁢

-0.51

POSITIVE LOGITS

 Nachdem

0.39

Policy

0.38

 Stande

0.34

tauscht

0.34

 Policy

0.34

 Sonn

0.34

fromnode

0.33

 encarga

0.33

EE

0.32

 purpose

0.32

Activations Density 0.025%

symbols, formatting, or structure related to document citations and references

No Comments

No Known Activations