INDEX

Explanations

references to specific dates and events related to personal experiences

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

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__res-sm_processed/2-res-sm

Prompts (Dashboard)

32,768 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

torch.float32

Hook Name

blocks.2.hook_resid_post

Hook Layer

Architecture

standard

Context Size

128

Dataset

EleutherAI/the_pile_deduplicated

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 respectively

-1.83

 separately

-1.74

 stratification

-1.54

 spectrum

-1.45

 strata

-1.44

 strat

-1.40

 distribution

-1.38

ncia

-1.38

oli

-1.37

 specifically

-1.37

POSITIVE LOGITS

ĥ½

3.21

±

2.97

»¿

2.72

ĩ

2.66

Ĭ

2.65

¤

2.65

ĥ

2.65

Ŀ

2.61

¦

2.58

ĵ

2.52

Activations Density 0.311%

references to specific dates and events related to personal experiences

No Comments

No Known Activations