INDEX

Explanations

expressions related to seeking answers or explanations

oai_token-act-pair · gpt-3.5-turbo

expressions of desire or intent

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

coded

-0.77

 hooked

-0.72

 clustered

-0.71

 fetish

-0.71

 nicknamed

-0.70

 famously

-0.70

pron

-0.69

 subclass

-0.68

 predicts

-0.67

 coined

-0.67

POSITIVE LOGITS

 justice

1.29

 reconciliation

1.19

 restitution

1.18

 peace

1.11

justice

1.10

Justice

1.10

 calm

1.09

 closure

1.06

 repentance

1.04

 healing

1.03

Activations Density 0.493%

expressions related to seeking answers or explanations

expressions of desire or intent

No Comments

No Known Activations