INDEX

Explanations

instances where the word 'Here' is used in the text

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

Spread

-0.73

odd

-0.71

Roll

-0.70

ÙĦ

-0.69

oooooooo

-0.68

TY

-0.67

CLOSE

-0.67

mic

-0.66

dule

-0.66

roll

-0.64

POSITIVE LOGITS

ãĥķãĤ©

0.72

ABE

0.66

aska

0.66

 differs

0.65

uca

0.65

 contrasts

0.64

 Recall

0.64

å£«

0.64

ayer

0.63

acha

0.63

Activations Density 0.060%

instances where the word 'Here' is used in the text

No Comments

No Known Activations