INDEX

Explanations

phrases emphasizing strong positive feelings or satisfaction

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "really" and its variations, indicating a focus on expressions of strong emphasis or intensity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

antry

-0.84

heid

-0.84

ansas

-0.78

icipated

-0.76

enary

-0.71

 Receiver

-0.70

wright

-0.66

tein

-0.66

 Purchase

-0.66

mare

-0.65

POSITIVE LOGITS

 darn

0.80

 appreciated

0.78

ignment

0.77

 distressed

0.73

 differentiated

0.71

 liked

0.71

FTWARE

0.70

 fucking

0.70

 fuckin

0.69

 helpful

0.68

Activations Density 0.051%

phrases emphasizing strong positive feelings or satisfaction

occurrences of the word "really" and its variations, indicating a focus on expressions of strong emphasis or intensity

No Comments

No Known Activations