INDEX

Explanations

occurrences of the word "Ter" at varying activation levels

oai_token-act-pair · gpt-3.5-turbo

mentions of specific terminology related to a product or service

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 strand

-0.77

 bike

-0.70

 beard

-0.68

 ears

-0.68

agog

-0.67

 mainline

-0.66

 pool

-0.66

 fools

-0.66

 pools

-0.65

 index

-0.65

POSITIVE LOGITS

Ter

3.91

TER

1.68

 Terms

1.48

terms

1.47

TER

1.41

Term

1.22

ter

1.19

Privacy

1.17

Termin

1.14

Rules

1.14

Activations Density 0.019%

occurrences of the word "Ter" at varying activation levels

mentions of specific terminology related to a product or service

No Comments

No Known Activations