INDEX

Explanations

forum discussion related metadata such as replies and post details

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.0.hook_resid_post_16384_anthropic

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

torch.float32

Hook Point

blocks.0.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

HuggingFaceFW/fineweb

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Godt

-0.82

 makro

-0.76

 effe

-0.76

hua

-0.75

 maksi

-0.75

UwU

-0.75

xxv

-0.74

uhr

-0.74

 Strukt

-0.74

 Diction

-0.71

POSITIVE LOGITS

Replies

3.17

<bos>

1.10

 reap

1.03

🕗

1.01

replies

0.88

Beskrivning

0.80

VYMaps

0.76

 Replies

0.75

 paradiso

0.74

Solución

0.70

Activations Density 0.278%

forum discussion related metadata such as replies and post details

No Comments

No Known Activations