INDEX

Explanations

terms related to different tiers or levels of classification

oai_token-act-pair · gpt-3.5-turbo

terms related to hierarchical classifications or levels

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Liberties

-0.78

mercial

-0.74

okia

-0.70

 Evening

-0.66

 Robo

-0.66

 Predators

-0.66

hur

-0.66

 Caption

-0.61

gotten

-0.61

vertisements

-0.60

POSITIVE LOGITS

 tier

1.18

 tiers

1.12

uple

0.96

Tier

0.93

 Tier

0.79

tier

0.78

 bracket

0.76

etting

0.69

 Canter

0.67

 veter

0.67

Activations Density 0.009%

terms related to different tiers or levels of classification

terms related to hierarchical classifications or levels

No Comments

No Known Activations