INDEX

Explanations

target words related to various activities and sports

oai_token-act-pair · gpt-3.5-turbo

gerunds and present participles related to various forms of movement or 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

ighed

-0.80

ifier

-0.79

ļéĨĴ

-0.77

witz

-0.76

slice

-0.76

angled

-0.70

zynski

-0.69

rez

-0.68

oping

-0.68

wait

-0.68

POSITIVE LOGITS

 prowess

1.14

 habits

1.12

 buddies

1.01

 styles

0.98

 techniques

0.96

redients

0.92

 opportunities

0.91

 etiquette

0.91

 sessions

0.91

 skills

0.90

Activations Density 0.228%

target words related to various activities and sports

gerunds and present participles related to various forms of movement or activity

No Comments

No Known Activations