INDEX

Explanations

mentions of remembering specific events or knowledge retrieval

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

oit

-0.78

osion

-0.75

ahime

-0.74

ourage

-0.74

avage

-0.72

assi

-0.71

tan

-0.71

aneers

-0.71

amen

-0.71

otomy

-0.71

POSITIVE LOGITS

remember

1.13

 vividly

1.11

 memories

1.04

 remembering

1.02

 nostalg

0.98

 remembered

0.97

 fond

0.96

 recalling

0.94

 Memories

0.91

 remembers

0.87

Activations Density 0.576%

mentions of remembering specific events or knowledge retrieval

No Comments

No Known Activations