INDEX

Explanations

number and symbol patterns mixed with words, potentially related to product specifications

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

New Auto-Interp

Configuration

jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Kiedy

-0.60

 bowiem

-0.56

 hornblende

-0.56

 bronco

-0.56

 chert

-0.54

 zirc

-0.53

 pelic

-0.53

 gaily

-0.52

 avanzado

-0.52

Wię

-0.52

POSITIVE LOGITS

 nessuna

0.80

 nessun

0.76

 cristina

0.73

 Più

0.73

 incess

0.72

 parlar

0.72

 allarg

0.71

 ritard

0.71

 apparti

0.70

 Queste

0.70

Activations Density 0.083%

number and symbol patterns mixed with words, potentially related to product specifications

No Comments

No Known Activations