INDEX

Explanations

specific acronyms "ABC" with different contexts and meanings

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "ABC."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Mi

-0.80

 turf

-0.71

 philosophy

-0.66

 lurking

-0.65

 clut

-0.65

 Slovakia

-0.65

 philosophical

-0.64

 Stur

-0.63

 stagnation

-0.63

Sed

-0.63

POSITIVE LOGITS

ABC

4.16

abc

2.63

ABC

2.34

NBC

1.87

CBS

1.81

CBC

1.75

BBC

1.50

CNN

1.46

FOX

1.39

NPR

1.34

Activations Density 0.012%

specific acronyms "ABC" with different contexts and meanings

occurrences of the term "ABC."

No Comments

No Known Activations