INDEX

Explanations

numerical values and scientific notation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res_post_32k-oai

Configuration

jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_6.pt

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

32,768

Data Type

torch.float32

Hook Name

blocks.6.hook_resid_post

Hook Layer

Architecture

standard

Context Size

Dataset

Skylion007/openwebtext

Activation Function

topk

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 crossings

-0.67

 pitfalls

-0.66

otin

-0.65

loss

-0.65

 etiquette

-0.62

 retaining

-0.61

killed

-0.60

 simplicity

-0.60

 ditch

-0.60

lings

-0.60

POSITIVE LOGITS

SPONSORED

0.83

 partName

0.80

 Fiat

0.73

itia

0.72

ItemImage

0.71

âĢİ

0.67

 à¨

0.65

[/

0.64

 æľ

0.62

âĢ¦âĢ¦

0.61

Activations Density 0.101%

numerical values and scientific notation

No Comments

No Known Activations