INDEX

Explanations

phrases related to falseness or imitation

oai_token-act-pair · gpt-3.5-turbo

terms indicating deceptive or false representations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

edin

-0.73

 Dynamics

-0.67

 sugg

-0.61

ials

-0.60

 screenings

-0.60

ells

-0.59

 drinks

-0.58

Õ

-0.58

rity

-0.58

 reservations

-0.57

POSITIVE LOGITS

judicial

0.84

legal

0.81

pas

0.81

icho

0.78

chal

0.75

cele

0.71

urrection

0.71

pas

0.71

reality

0.71

éĹ

0.69

Activations Density 0.104%

phrases related to falseness or imitation

terms indicating deceptive or false representations

No Comments

No Known Activations