INDEX

Explanations

phrases related to news headlines or current events

oai_token-act-pair · gpt-3.5-turbo

significant predictive statements about potential events or outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

_.

-0.81

example

-0.77

-0.76

 assum

-0.76

them

-0.76

 thereof

-0.72

 Niet

-0.72

âĶĢâĶĢâĶĢâĶĢ

-0.71

ãĢĤ

-0.71

thing

-0.70

POSITIVE LOGITS

 honoured

0.76

 watchdog

0.75

 declass

0.75

 Wednesday

0.75

 Thursday

0.72

 unveiled

0.70

 cybersecurity

0.68

 apologised

0.68

 renewed

0.68

 Tuesday

0.68

Activations Density 0.523%

phrases related to news headlines or current events

significant predictive statements about potential events or outcomes

No Comments

No Known Activations