INDEX

Explanations

references to potentially harmful or negative elements in narratives

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

-0.46

-0.45

-0.43

-0.42

-0.41

-0.40

<eos>

-0.40

be

-0.39

POSITIVE LOGITS

 Monfieur

1.32

 myſelf

1.29

 itſelf

1.20

 themſelves

1.18

 himſelf

1.16

 ſtate

1.16

ſelf

1.10

 nakalista

1.10

 Majefty

1.10

 ſta

1.08

Activations Density 0.315%

references to potentially harmful or negative elements in narratives

No Comments

No Known Activations