INDEX

Explanations

phrases related to weighing options or making choices based on trade-offs and consequences

oai_token-act-pair · gpt-3.5-turbo

concepts related to trade-offs and dualities

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gur

-0.78

leck

-0.69

 screened

-0.68

 slides

-0.68

 rolls

-0.68

 carts

-0.68

FP

-0.68

 desks

-0.66

intend

-0.66

visors

-0.65

POSITIVE LOGITS

 inherent

0.90

 implications

0.88

 equation

0.83

 implication

0.83

 dich

0.81

why

0.81

 paradox

0.80

terness

0.78

 tang

0.78

 irony

0.78

Activations Density 0.410%

phrases related to weighing options or making choices based on trade-offs and consequences

concepts related to trade-offs and dualities

No Comments

No Known Activations