INDEX

Explanations

phrases related to permission or allowance

oai_token-act-pair · gpt-3.5-turbo

phrases that suggest enabling or allowing specific actions or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Notting

-0.77

ĸļ

-0.74

Sax

-0.72

alon

-0.69

ernel

-0.66

edin

-0.65

onday

-0.64

 Blitz

-0.63

herer

-0.63

æ©Ł

-0.63

POSITIVE LOGITS

 impunity

0.72

 access

0.71

 lapse

0.70

 lookup

0.67

undai

0.66

 safely

0.66

 unchecked

0.65

 carve

0.65

 unrestricted

0.65

 unlimited

0.65

Activations Density 0.349%

phrases related to permission or allowance

phrases that suggest enabling or allowing specific actions or conditions

No Comments

No Known Activations