INDEX

Explanations

words related to the cockpit of an aircraft

oai_token-act-pair · gpt-3.5-turbo

references to aircraft and cockpit terminology

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

 Flavoring

-1.14

å¥³

-1.10

roots

-0.84

chin

-0.80

hens

-0.79

Avg

-0.79

¬¼

-0.78

 Athlet

-0.77

arians

-0.76

 Vegan

-0.75

POSITIVE LOGITS

selage

0.93

 dispatcher

0.88

 rotor

0.88

 subsystem

0.88

ilot

0.87

 regulator

0.86

 navigation

0.85

 compartment

0.82

 cockpit

0.79

 malfunction

0.77

Activations Density 0.127%

words related to the cockpit of an aircraft

references to aircraft and cockpit terminology

No Comments

No Known Activations