INDEX

Explanations

names of people, particularly their familial relationships and affiliations

oai_token-act-pair · gpt-3.5-turbo

names and relationships among individuals

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rums

-0.63

 manually

-0.63

 unprepared

-0.62

ript

-0.62

 excuses

-0.61

limits

-0.61

Against

-0.60

 pointless

-0.59

 notifications

-0.58

 readiness

-0.57

POSITIVE LOGITS

 fame

1.06

Sr

0.93

 founder

0.88

 whom

0.85

 patriarch

0.85

 ancestor

0.84

 deceased

0.81

 dynasty

0.80

who

0.78

enegger

0.77

Activations Density 0.328%

names of people, particularly their familial relationships and affiliations

names and relationships among individuals

No Comments

No Known Activations