INDEX

Explanations

words related to administrative tasks and processes

oai_token-act-pair · gpt-3.5-turbo

specific terms related to scheduling and grading systems in various contexts

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

ngth

-0.84

ived

-0.77

ãĥĩãĤ£

-0.76

essim

-0.73

ogn

-0.73

ieving

-0.73

ethe

-0.72

icism

-0.72

ewitness

-0.71

playing

-0.71

POSITIVE LOGITS

 methods

1.03

 techniques

1.03

 mechanism

1.02

 mechanisms

1.00

 procedures

1.00

 algorithms

0.96

 practices

0.95

 conventions

0.94

 apparatus

0.93

 schemes

0.92

Activations Density 0.218%

words related to administrative tasks and processes

specific terms related to scheduling and grading systems in various contexts

No Comments

No Known Activations