INDEX

Explanations

terms related to orientation and positioning

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_post_32k-oai

Configuration

jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

32,768

Data Type

torch.float32

Hook Name

blocks.8.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

enegger

-0.88

soDeliveryDate

-0.84

channelAvailability

-0.73

icas

-0.73

isSpecialOrderable

-0.69

 Mellon

-0.68

tons

-0.66

ãĤ¨ãĥ«

-0.65

utical

-0.65

hire

-0.65

POSITIVE LOGITS

eering

1.33

ational

0.90

¥µ

0.82

ated

0.79

 toward

0.79

irection

0.79

£

0.78

Īè

0.78

 orientation

0.73

 shift

0.73

Activations Density 0.091%

terms related to orientation and positioning

No Comments

No Known Activations