INDEX

Explanations

names of individuals

oai_token-act-pair · gpt-3.5-turbo

proper names, specifically individuals' names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ocial

-0.91

soc

-0.89

leaders

-0.84

vag

-0.83

flies

-0.82

¥µ

-0.82

subject

-0.80

services

-0.80

hospital

-0.79

safety

-0.79

POSITIVE LOGITS

 Snyder

0.97

 Rogers

0.97

 Martinez

0.96

 Rudd

0.94

 Reeves

0.94

 Fernandez

0.94

 Manuel

0.93

 Murray

0.93

 Morrison

0.92

 Torres

0.92

Activations Density 0.123%

names of individuals

proper names, specifically individuals' names

No Comments

No Known Activations