INDEX

Explanations

phrases indicating actions or necessities

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

ahime

-0.73

âĦ¢:

-0.72

 prominently

-0.69

 esche

-0.64

*:

-0.63

 incorporates

-0.62

Specifically

-0.61

iqueness

-0.60

hig

-0.60

 showcased

-0.60

POSITIVE LOGITS

 anymore

0.82

 except

0.71

ope

0.68

ARE

0.66

Else

0.65

done

0.64

now

0.63

empty

0.60

chat

0.60

else

0.60

Activations Density 0.565%

phrases indicating actions or necessities

No Comments

No Known Activations