INDEX

Explanations

negative forms of verbs and inquiries about knowledge

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

antMatchers

-0.83

XmlAccessType

-0.72

stateMutability

-0.64

 trône

-0.64

いる

-0.60

ormais

-0.58

]").

-0.58

onomía

-0.57

Projection

-0.57

 publiques

-0.56

POSITIVE LOGITS

did

1.20

did

1.16

Did

1.13

DID

1.05

DID

1.05

Did

1.02

 didn

0.99

 didnt

0.98

 Didn

0.91

Twas

0.91

Activations Density 0.207%

negative forms of verbs and inquiries about knowledge

No Comments

No Known Activations