INDEX

Explanations

words related to wire and wire-related activities

oai_token-act-pair · gpt-3.5-turbo

references to "wire" or related terminology, often in legal or technical 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

ivia

-0.86

eston

-0.72

 Scotia

-0.68

anchester

-0.64

uca

-0.64

BIP

-0.63

ļé

-0.62

ãħĭ

-0.62

iterranean

-0.61

 Flav

-0.61

POSITIVE LOGITS

lessly

1.25

tap

1.14

frame

1.07

fram

1.03

wire

0.95

haired

0.93

 wire

0.91

frames

0.89

 mesh

0.88

 wires

0.86

Activations Density 0.011%

words related to wire and wire-related activities

references to "wire" or related terminology, often in legal or technical contexts

No Comments

No Known Activations