INDEX

Explanations

time-related phrases, particularly durations like days, weeks, and months

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Reviewer

-0.78

ahime

-0.70

acus

-0.70

ole

-0.67

eret

-0.65

orescence

-0.64

olas

-0.62

psons

-0.61

inite

-0.58

iots

-0.57

POSITIVE LOGITS

ago

1.21

 ahead

1.04

Ago

0.94

shy

0.93

 apart

0.89

 kilometers

0.88

 weeks

0.87

 away

0.87

 kilometres

0.84

 miles

0.82

Activations Density 0.998%

time-related phrases, particularly durations like days, weeks, and months

No Comments

No Known Activations