INDEX

Explanations

text related to scientific theories and hypotheses

oai_token-act-pair · gpt-3.5-turbo

concepts related to scientific reasoning and hypothesis formation

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

audi

-0.76

©¶æ¥µ

-0.71

cussion

-0.68

Ħ¢

-0.67

phal

-0.61

devices

-0.61

 laptops

-0.60

ashtra

-0.59

BuyableInstoreAndOnline

-0.58

Bridge

-0.57

POSITIVE LOGITS

 somebody

0.93

 someone

0.86

someone

0.81

 others

0.78

 instantly

0.76

 describ

0.76

 ordinarily

0.73

terday

0.72

 another

0.69

 magically

0.69

Activations Density 0.807%

text related to scientific theories and hypotheses

concepts related to scientific reasoning and hypothesis formation

No Comments

No Known Activations