INDEX

Explanations

time-related expressions like durations and timeframes

oai_token-act-pair · gpt-3.5-turbo

temporal references, particularly relating to past and future durations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Released

-0.70

éŃĶ

-0.70

 Released

-0.67

emale

-0.65

Synopsis

-0.65

ãĥ´

-0.64

elist

-0.60

vale

-0.58

edIn

-0.56

icipated

-0.56

POSITIVE LOGITS

or

1.23

 guys

1.19

 maybe

1.15

 fuckin

1.04

['

1.02

?"

1.02

 bucks

0.99

-,

0.94

.""

0.89

Activations Density 0.176%

time-related expressions like durations and timeframes

temporal references, particularly relating to past and future durations

No Comments

No Known Activations