INDEX

Explanations

words related to wires or wire-related activities

oai_token-act-pair · gpt-3.5-turbo

references to "wire."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 McDonnell

-0.69

 Dunk

-0.63

 Kaepernick

-0.62

 Modest

-0.62

 Vall

-0.62

 Camb

-0.62

 Graham

-0.61

 Scand

-0.61

San

-0.61

Instance

-0.61

POSITIVE LOGITS

 wire

1.32

 wires

1.09

wire

1.02

 Wire

1.02

kefeller

1.01

 wiring

0.89

Wire

0.84

fram

0.84

oshenko

0.84

haired

0.83

Activations Density 0.005%

words related to wires or wire-related activities

references to "wire."

No Comments

No Known Activations