INDEX

Explanations

words related to magnetic fields and forces

oai_token-act-pair · gpt-3.5-turbo

references to magnetic fields and their effects

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

 Homeless

-0.77

 Hispanic

-0.75

 McCann

-0.73

pkg

-0.70

Elm

-0.69

 Nicarag

-0.68

 Latino

-0.67

 Faces

-0.67

 Cinnamon

-0.67

 DACA

-0.66

POSITIVE LOGITS

 exerted

1.08

 coefficient

1.07

 currents

1.02

 vibrations

1.00

 induced

0.91

 vibration

0.91

 amplitude

0.90

 resonance

0.90

 coefficients

0.89

 fluctuations

0.89

Activations Density 0.187%

words related to magnetic fields and forces

references to magnetic fields and their effects

No Comments

No Known Activations