INDEX

Explanations

terms related to cause and effect, often focusing on unintended consequences and cascading effects

oai_token-act-pair · gpt-3.5-turbo

themes related to ripple effects and unintended consequences

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

ighters

-0.80

unte

-0.73

enhagen

-0.71

fighters

-0.69

akov

-0.68

uits

-0.67

fighter

-0.66

packing

-0.65

 fries

-0.65

estine

-0.65

POSITIVE LOGITS

 ripple

1.14

å¥³

0.99

Effects

0.97

 effect

0.91

Effect

0.89

 Effect

0.85

romeda

0.82

effects

0.82

effect

0.81

 casc

0.78

Activations Density 0.059%

terms related to cause and effect, often focusing on unintended consequences and cascading effects

themes related to ripple effects and unintended consequences

No Comments

No Known Activations