INDEX

Explanations

direct connections or relationships between objects or individuals mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

strong associations with the word "direct"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

abies

-0.96

anners

-0.80

 Mell

-0.74

tex

-0.73

nesota

-0.72

mbuds

-0.71

gerald

-0.71

 Kirin

-0.70

 Fitzgerald

-0.70

kees

-0.69

POSITIVE LOGITS

 direct

0.93

 sunlight

0.87

eur

0.86

 injection

0.76

oire

0.76

 confrontation

0.75

 causation

0.74

 indirect

0.73

 contact

0.73

 blast

0.72

Activations Density 0.015%

direct connections or relationships between objects or individuals mentioned in the text

strong associations with the word "direct"

No Comments

No Known Activations