INDEX

Explanations

terms related to oaths, pledges, and commitments

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

nesota

-0.76

ogo

-0.65

 guiActiveUnfocused

-0.65

Stage

-0.64

DragonMagazine

-0.64

markets

-0.61

ption

-0.59

Ļ½

-0.58

 Closed

-0.58

CDC

-0.56

POSITIVE LOGITS

 oath

1.35

 allegiance

1.15

 sworn

1.08

 swearing

1.05

 swore

0.98

 swear

0.96

vow

0.96

breaker

0.94

 solemn

0.91

 Oath

0.88

Activations Density 0.191%

terms related to oaths, pledges, and commitments

No Comments

No Known Activations