INDEX

Explanations

mentions of endings or conclusions

oai_token-act-pair · gpt-3.5-turbo

suffixes and endings related to various processes or categories

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 showc

-0.83

 carbohyd

-0.82

amous

-0.77

 tremend

-0.77

 gobl

-0.72

 veget

-0.70

 thor

-0.70

 disarm

-0.69

 satell

-0.68

 toget

-0.68

POSITIVE LOGITS

 Association

1.19

 Authority

1.08

 Club

1.06

 Suite

1.02

 Service

1.00

 Tracker

0.99

 Coordinator

0.99

 Committee

0.98

 Foundation

0.98

 Award

0.98

Activations Density 0.305%

mentions of endings or conclusions

suffixes and endings related to various processes or categories

No Comments

No Known Activations