INDEX

Explanations

negative references and denials

oai_token-act-pair · gpt-4o-mini

New Auto-Interp

Configuration

chanind/gemma-2-2b-batch-topk-matryoshka-saes-w-32k-l0-40/standard/blocks.12.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Architecture

jumprelu

Context Size

1,024

Dataset

chanind/pile-uncopyrighted-gemma-1024-abbrv-1B

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lock

-0.53

 lock

-0.51

LOCK

-0.47

Lock

-0.46

arki

-0.45

 Ball

-0.45

resave

-0.44

 Parke

-0.44

Du

-0.44

mittedly

-0.44

POSITIVE LOGITS

AndEndTag

0.75

IntoConstraints

0.75

 surla

0.69

AccessorTable

0.69

enumi

0.61

UnusedPrivate

0.60

tagHelperRunner

0.60

}}</

0.60

 jäm

0.59

WindowConstants

0.59

Activations Density 0.184%

negative references and denials

No Comments

No Known Activations