INDEX

Explanations

phrases indicating possession or ownership

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res_post_32k-oai

Configuration

jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

32,768

Data Type

torch.float32

Hook Name

blocks.10.hook_resid_post

Hook Layer

Architecture

standard

Context Size

Dataset

Skylion007/openwebtext

Activation Function

topk

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 mastered

-0.71

 writes

-0.69

 WRITE

-0.68

 dresses

-0.68

 wrote

-0.68

 typed

-0.68

 knows

-0.67

 regrets

-0.66

 rapp

-0.66

 hates

-0.66

POSITIVE LOGITS

 resulted

0.89

 outraged

0.87

 prompted

0.86

 ramifications

0.86

 alarmed

0.86

 implications

0.84

 sparked

0.82

iatrics

0.80

 baffled

0.80

igor

0.78

Activations Density 0.570%

phrases indicating possession or ownership

No Comments

No Known Activations