INDEX

Explanations

names of people or characters in a specific context

oai_token-act-pair · gpt-3.5-turbo

names of people or entities involved in various narratives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

mining

-0.55

PDATED

-0.54

phas

-0.53

impact

-0.53

mble

-0.52

 reclaimed

-0.51

msg

-0.51

paced

-0.51

 binds

-0.51

Manufact

-0.51

POSITIVE LOGITS

 backstage

0.96

Jr

0.95

 during

0.94

Sr

0.83

 afterward

0.82

 shortly

0.82

 onstage

0.81

 beforehand

0.80

 sometime

0.78

 afterwards

0.78

Activations Density 0.272%

names of people or characters in a specific context

names of people or entities involved in various narratives

No Comments

No Known Activations