INDEX

Explanations

dates or chronological events

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate temporal sequences or events

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

ranged

-0.88

continue

-0.72

atom

-0.67

objects

-0.67

incre

-0.66

wart

-0.66

imentary

-0.66

options

-0.64

iary

-0.64

cel

-0.64

POSITIVE LOGITS

 dusk

0.91

 ironic

0.88

 bitters

0.86

 dawn

0.83

htaking

0.80

 raining

0.77

 twilight

0.77

 surreal

0.77

 thrilling

0.77

TIME

0.76

Activations Density 0.187%

dates or chronological events

phrases that indicate temporal sequences or events

No Comments

No Known Activations