INDEX

Explanations

phrases related to discussions or consultations

oai_token-act-pair · gpt-3.5-turbo

phrases associated with discussions and consultations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ãĥĺ

-0.75

iru

-0.73

SPONSORED

-0.71

edes

-0.68

 Coverage

-0.67

escape

-0.67

 Falling

-0.63

 Citation

-0.63

apo

-0.61

 impunity

-0.61

POSITIVE LOGITS

 internally

1.08

 brainstorm

0.99

actionDate

0.98

 beforehand

0.98

 consultations

0.93

 consultation

0.93

 briefed

0.90

 consulted

0.90

 discussions

0.85

 mediation

0.81

Activations Density 0.449%

phrases related to discussions or consultations

phrases associated with discussions and consultations

No Comments

No Known Activations