INDEX

Explanations

instances of objects being filled or equipped with something, such as a van being loaded with people or a gun being loaded with bullets

oai_token-act-pair · gpt-3.5-turbo

references to the term "loaded" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

agara

-0.78

fax

-0.74

cko

-0.74

study

-0.71

recogn

-0.70

Ĥİ

-0.67

kos

-0.67

ateur

-0.67

perty

-0.66

chal

-0.65

POSITIVE LOGITS

 loads

0.93

loading

0.93

 loaded

0.86

 baskets

0.83

loaded

0.80

raft

0.78

 crates

0.75

 load

0.74

loads

0.73

 loading

0.71

Activations Density 0.012%

instances of objects being filled or equipped with something, such as a van being loaded with people or a gun being loaded with bullets

references to the term "loaded" in various contexts

No Comments

No Known Activations