INDEX

Explanations

names of people, particularly repeating ones, likely indicating a focus on a specific individual

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 impractica

-0.88

 liberality

-0.87

 McLaugh

-0.85

 ecru

-0.82

 unlaw

-0.77

 McInt

-0.72

 nutella

-0.69

 Ename

-0.68

 Rine

-0.66

 Heywood

-0.66

POSITIVE LOGITS

 Davis

1.58

Davis

1.40

 DAVIS

1.15

davis

1.12

 davis

0.93

 Davies

0.78

DAV

0.73

datasets

0.62

Dataset

0.60

 dataset

0.58

Activations Density 0.109%

names of people, particularly repeating ones, likely indicating a focus on a specific individual

No Comments

No Known Activations