INDEX

Explanations

locations or areas that are associated with specific activities or properties

oai_token-act-pair · gpt-3.5-turbo

terms related to significant physical locations or centers of activity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Overt

-0.55

 warr

-0.54

Austral

-0.54

 Seah

-0.52

INGS

-0.52

INC

-0.52

REE

-0.52

Auth

-0.51

 motions

-0.51

 PROG

-0.50

POSITIVE LOGITS

for

1.10

 conducive

0.87

 ripe

0.85

 destination

0.82

hub

0.81

for

0.80

 frequ

0.79

 refuge

0.78

 venue

0.77

 where

0.74

Activations Density 0.363%

locations or areas that are associated with specific activities or properties

terms related to significant physical locations or centers of activity

No Comments

No Known Activations