INDEX

Explanations

proper nouns related to names, possibly of people or places

oai_token-act-pair · gpt-3.5-turbo

names and references associated with specific individuals or characters

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Chem

-0.70

AMA

-0.68

 Participant

-0.67

 Hannah

-0.66

atches

-0.66

 gestation

-0.65

Food

-0.63

 Pocket

-0.63

 Polo

-0.61

 Bastard

-0.60

POSITIVE LOGITS

doms

0.91

riks

0.90

Reloaded

0.81

hyde

0.79

awa

0.77

nil

0.74

":"/

0.74

ting

0.74

nery

0.74

igation

0.73

Activations Density 0.030%

proper nouns related to names, possibly of people or places

names and references associated with specific individuals or characters

No Comments

No Known Activations