INDEX

Explanations

mentions of bandwidth and throughput metrics

oai_token-act-pair · gpt-3.5-turbo

terms related to network bandwidth and throughput

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 Kubrick

-0.73

oan

-0.72

neys

-0.69

inse

-0.69

 Martinez

-0.68

tail

-0.67

ten

-0.66

 Garfield

-0.65

NEY

-0.64

 Bain

-0.64

POSITIVE LOGITS

 bandwidth

1.39

width

0.92

 throughput

0.89

 optimizations

0.84

 utilization

0.81

 allocations

0.80

velength

0.79

 aperture

0.79

 connectivity

0.79

 latency

0.79

Activations Density 0.009%

mentions of bandwidth and throughput metrics

terms related to network bandwidth and throughput

No Known Activations