INDEX

Explanations

terms related to computer networks

oai_token-act-pair · gpt-3.5-turbo

references to networks, particularly in technical or systemic contexts

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

 Lama

-0.74

orical

-0.68

ruciating

-0.68

mond

-0.67

erenn

-0.66

Ë

-0.64

 Fraz

-0.63

ugh

-0.63

juries

-0.63

gg

-0.62

POSITIVE LOGITS

 networks

0.95

 connectivity

0.88

 network

0.88

 Tycoon

0.84

 Networks

0.83

 routers

0.82

 connection

0.77

 connections

0.75

 administrator

0.74

ed

0.74

Activations Density 0.027%

terms related to computer networks

references to networks, particularly in technical or systemic contexts

No Comments

No Known Activations