INDEX

Explanations

words related to technology, specifically programming and usage of specific technologies

oai_token-act-pair · gpt-3.5-turbo

words indicating location, direction, or association

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

SHIP

-0.64

 compared

-0.62

 pitched

-0.60

arthed

-0.60

oulos

-0.60

 wielded

-0.58

âĶĢ

-0.58

 authored

-0.56

 Sorceress

-0.56

 accessed

-0.55

POSITIVE LOGITS

pless

1.08

effic

0.76

animate

0.76

oreal

0.74

roads

0.73

ordinate

0.73

odes

0.70

arus

0.68

onement

0.67

authent

0.66

Activations Density 0.477%

words related to technology, specifically programming and usage of specific technologies

words indicating location, direction, or association

No Comments

No Known Activations