INDEX

Explanations

phrases related to giving approval or permission

oai_token-act-pair · gpt-3.5-turbo

terms related to approval or permission, particularly in the context of "green light" signals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

sonian

-0.80

oln

-0.72

halla

-0.72

================================================================

-0.71

apter

-0.71

 Mehran

-0.70

ocal

-0.66

uum

-0.66

IRD

-0.66

ersive

-0.65

POSITIVE LOGITS

lights

0.95

lighting

0.87

 bulb

0.80

light

0.77

edly

0.76

ed

0.76

beam

0.74

osate

0.73

 bulbs

0.72

lit

0.72

Activations Density 0.016%

phrases related to giving approval or permission

terms related to approval or permission, particularly in the context of "green light" signals

No Comments

No Known Activations