INDEX

Explanations

phrases related to technical operations or instructions

oai_token-act-pair · gpt-3.5-turbo

terms related to functionality and operational status

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

ylum

-0.78

mere

-0.75

esan

-0.72

 Hague

-0.69

Ń·

-0.68

Ware

-0.67

iasis

-0.67

vier

-0.67

AppData

-0.66

cott

-0.66

POSITIVE LOGITS

 throttle

0.88

 ignition

0.82

 levers

0.81

 actu

0.80

 functioning

0.80

 valve

0.78

 amps

0.77

 cycle

0.76

 valves

0.75

 regulator

0.74

Activations Density 0.298%

phrases related to technical operations or instructions

terms related to functionality and operational status

No Comments

No Known Activations