INDEX

Explanations

questions and statements regarding expectations, learning, and accountability

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hement

-0.70

quad

-0.67

scill

-0.60

clerosis

-0.60

abad

-0.59

ioxide

-0.59

ric

-0.59

rf

-0.58

rontal

-0.57

 aired

-0.57

POSITIVE LOGITS

 accomplished

0.99

 accomplish

0.96

 Learned

0.84

 Doing

0.81

 doing

0.81

 done

0.79

 learnt

0.78

 wrought

0.78

 entail

0.76

 achieve

0.74

Activations Density 0.143%

questions and statements regarding expectations, learning, and accountability

No Comments

No Known Activations