INDEX

Explanations

terms related to power sources

oai_token-act-pair · gpt-3.5-turbo

references to power sources or powering mechanisms

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

esan

-0.79

ordan

-0.75

ormon

-0.73

istani

-0.71

ãģĵ

-0.70

ormons

-0.70

algia

-0.69

 ç¥ŀ

-0.69

icho

-0.67

eeper

-0.64

POSITIVE LOGITS

 powered

0.81

whe

0.76

PDATE

0.76

 powering

0.76

wagen

0.74

 turbines

0.70

powered

0.70

 Powered

0.70

 wheels

0.67

 propulsion

0.67

Activations Density 0.043%

terms related to power sources

references to power sources or powering mechanisms

No Comments

No Known Activations